use of neutron beams for low and medium flux …iaea-tecdoc-836 use of neutron beams for low and...

IAEA-TECDOC-836

Use of neutron beams forlow and medium flux

research reactors:R&D programmes in

materials scienceReport of an Advisory Group meeting

held in Vienna, 29 M arch-1 April 1993

INTERNATIONAL ATOMIC ENERGY AGENCY

The IAEA does not normally maintain stocks of reports in this series.However, microfiche copies of these reports can be obtained from

IN IS ClearinghouseInternational Atomic Energy AgencyWagramerstrasse 5P.O. Box 100A-1400 Vienna, Austria

Orders should be accompanied by prepayment of Austrian Schillings 100,in the form of a cheque or in the form of IAEA microfiche service couponswhich may be ordered separately from the INIS Clearinghouse.

The originating Section of this publication in the IAEA was:Physics Section

International Atomic Energy AgencyWagramerstrasse 5

P.O. Box 100A-1400 Vienna, Austria

USE OF NEUTRON BEAMS FOR LOW AND MEDIUM FLUX RESEARCH REACTORS:R&D PROGRAMMES IN MATERIALS SCIENCE

IAEA, VIENNA, 1995IAEA-TECDOC-836

ISSN 1011-4289

©IAEA, 1995

Printed by the IAEA in AustriaOctober 1995

FOREWORD

Research reactors have been playing an important role in the development of scientificand technological infrastructure and in training of manpower for the introduction of nuclearpower in many countries. Currently, there are 284 operational research reactors in the world,including 88 research reactors hi 39 developing countries; the number of reactors indeveloping countries is increasing as more countries embark on programmes in nuclearscience and technology. However, full utilization of these facilities for fundamental andapplied research has seldom been achieved. In particular, the utilization of beam ports hasbeen quite low.

Neutron beam based research is regarded as one of the most important programme areasthat can be implemented even at low and medium flux reactors. The range of activitiespossible in this field is so wide that it is generally feasible to define R&D programmes to suitdifferent needs and conditions. It is, therefore, important and of direct benefit to identifyeffective ways to improve utilization of beam tubes and of the research reactors. To this end,the International Atomic Energy Agency organized two meetings during 1993. First, anAdvisory Group Meeting (AGM) on Use of Research Reactors for Solid State Studies andsecond, a Technical Committee Meeting (TCM) on Use of Research Reactor Neutron Beamsfor Radiography and Materials Characterization. The goal of the AGM was to review andprovide recommendations on neutron scattering techniques, while the TCM focused onneutron radiography and other simple methods and applications.

The present report is the result of the Advisory Group Meeting held during 29 March- 1 April 1993 in Vienna with subsequent contributions from the experts. The PhysicsSection of the Department of Research and Isotopes was responsible for the co-ordination andcompilation of the report.

The report is intended to provide guidelines to research reactor owners and operatorsfor promoting and developing neutron beam based research programmes for solid statestudies using neutron scattering techniques. It is expected to benefit ongoing facilities andprogrammes by encouraging use of unproved techniques for detection, signal acquisition,signal processing, etc. and new programmes by assisting in the selection of appropriateequipment, instrument design and research plans.

EDITORIAL NOTE

In preparing this publication for press, staff, of the IAEA have made up the pages from theoriginal manuscripts as submitted by the authors. The views expressed do not necessarily reflect thoseof the governments of the nominating Member States or of the nominating organizations.

Throughout the text names of Member States are retained as they were when the text wascompiled.

The use of particular designations of countries or territories does not imply any judgement bythe publisher, the IAEA, as to the legal status of such countries or territories, of their authorities andinstitutions or of the delimitation of their boundaries.

The mention of names of specific companies or products (whether or not indicated as registered)does not imply any intention to infringe proprietary rights, nor should it be construed as anendorsement or recommendation on the part of the IAEA.

The authors are responsible for having obtained the necessary permission for the IAEA toreproduce, translate or use material from sources already protected by copyrights.

CONTENTS

1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2. FRAMEWORK FOR SUSTAINABLE R&D . . . . . . . . . . . . . . . . . . . . . . . 7

2.1. Importance of neutron scattering techniques . . . . . . . . . . . . . . . . . . . . 72.2. Minimum size of research groups . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.3. Infrastructural facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.4. Essential investment needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.5. Reverse mobility of scientists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.6. Interdisciplinary collaboration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3. FUTURE RESEARCH PROGRAMMES . . . . . . . . . . . . . . . . . . . . . . . . . 93.1. Magnetic compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103.2. Magnetically disordered systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 103.3. Magnetic thin films and superlattices . . . . . . . . . . . . . . . . . . . . . . . . . 113.4. Ceramic (high T,.) superconductors . . . . . . . . . . . . . . . . . . . . . . . . . . 113.5. Texture and residual stress in fabricated materials . . . . . . . . . . . . . . . . . 123.6. Zeolites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.7. Petrogenic materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.8. Minerals, polymers and amorphous systems . . . . . . . . . . . . . . . . . . . . 12

4. METHODS AND INSTRUMENTATION . . . . . . . . . . . . . . . . . . . . . . . . 13

4.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134.2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134.3. Techniques and instrumentation details . . . . . . . . . . . . . . . . . . . . . . . 154.4. Data acquisition and control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

5. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

ANNEX: PAPERS PRESENTED AT THE ADVISORY GROUP MEETING

Fundamental and applied neutron research at the 250 kW TRIGA reactor Vienna ... 29G. Badurek

Use of research reactors for solid state physics studies at Bhabha AtomicResearch Centre, India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41K.R. Rao

Solid state physics around Pakistan Research Reactor at PINSTECH . . . . . . . . . . . 55N.M. Butt, J. Bashir, M. Nasir Khan

Framework for a sustainable development of neutron beam work in the smallerresearch reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81F. G. Carvalho, P.M.A. Margaça

Basic research in solid state physics using a small research reactor . . . . . . . . . . . . 91V. Dimic

The new Swiss spallation neutron source SINQ and its planned day-one instruments —Status as of spring 1993 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97G.S. Bauer

Neutron scattering research with low to medium flux reactors . . . . . . . . . . . . . . . 1 1 5J.J. Khyne

The ORPHEE versatile low-cost multiprocessor system for data acquisition andcontrol of neutron spectrometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125G. Koskas

List of Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

1. INTRODUCTION

Worldwide there are 284 operable nuclear research reactors of which 88 are located in390 developing countries. These research reactors and the laboratories around them representa significant research potential and a source of highly qualified and trained manpower notonly for activities related to nuclear fields but also for several conventional technologies. Theacquisition of a nuclear research reactor, in a developing country, is generally justified onthe basis that it is a necessary tool for:

• Fundamental and applied research in various disciplines;• Training of young scientists in nuclear sciences;• Production of radioisotopes;• Supporting activities related to nuclear power including operator training;• Engineering experiments.

These objectives have been achieved with different degrees of success in differentcountries. Among several factors that contribute toward underutilization of these facilities,the most obvious ones are: inappropriate research programmes, lack of adequate equipmentand laboratories, relatively low neutron fluxes, non-availability of sufficiently trainedmanpower. In particular, the utilization of neutron beam ports remains very low. This is inspite of the fact that neutron beam based research is one of the most important programmeareas which can be implemented, even at low and medium flux reactors. The range ofactivities possible in this area is so wide that it is in general possible to define and carry outR&D programmes to suit specific needs and conditions.

This report covers various aspects of neutron beam research applicable to materialsscience. It specifies minimal requirements concerning infrastructures, personnel, funding, etc.as these conditions must coexist for a viable research programme over a period of severalyears. A brief overview of R&D programmes on condensed matter physics with the requiredexperimental methods, instruments and techniques is presented. Depending upon localconditions, the report is intended to serve as a technical document to guide, initiate, review,upgrade or support programmes in neutron beam based R&D. Topics like production ofradioisotopes, irradiation, neutron radiography, nuclear physics, etc. are not covered. Relatedsubjects on neutron radiography, irradiation, etc. are covered separately in IAEA-TECDOC-837. The papers presented by the participants are included in the Annex.

The report will benefit the ongoing facilities and programmes by encouraging the useof improved techniques for detection, signal acquisition, signal processing, etc. and the newprogrammes in the selection of appropriate equipment, instrument design, and planning ofresearch. It will also be useful for upgrading and modernization of the aging experimentalfacilities.

2. FRAMEWORK FOR SUSTAINABLE R&D

2.1. IMPORTANCE OF NEUTRON SCATTERING TECHNIQUES

Neutron beam based R&D is one of the most versatile activities in scientific andtechnical applications, particularly for relating the bulk properties to microscopic structure(and dynamics) for the development of new materials and techniques. The neutron scatteringtechniques have proved very fruitful for investigating structural properties of complexmaterials and in obtaining information on the relationships between structure, properties,

performance and the effects of various processes. As a tool for the characterization ofmaterials, low-energy neutrons offer several unique advantages over other probes such asprotons, electrons or X rays. They are also useful in quality assurance of manufacturedproducts, optimization of process parameters and also in the study of fatigue and stressphenomena. These are particularly important aspects for developing countries who desire todevelop or assist manufacturing industries and to increase added value to indigenous rawmaterials. Since neutrons are chargeless particles, capable of penetrating bulk samples, theydo not suffer from large form factor fall-offs at high momentum transfers as is the case withX ray. Neutrons have a magnetic moment and therefore provide information on magneticstate of samples although in this case a form factor comes into play. The energies of neutronsmatch well with those of excitation energies in condensed matter and are therefore extremelyuseful to study the nature of excitation spectra across the entire Brillouin zone in solids.

2.2. MINIMUM SIZE OF RESEARCH GROUPS

Fundamental requirement for initiating a viable R&D work in neutron beam researchat any research center is the availability of a minimum number of scientists and technicianswith appropriate specialization and skills. The smallest 'critical' group should comprise 3qualified scientists (at least one Ph.D.) and 2 technicians. This group should be encouragedto have reasonable contacts with local and foreign scientists in their fields of work. If at anystage, an experienced scientist must be moved out from this small group, the replacementshould be provided as soon as possible and continuity should be ensured. Depending uponthe qualification and experience of the newcomer, it will be necessary to allow for areasonable period of overlap. In any case the group should be maintained by retaining at leastone of the experienced scientists. Such a small group can initiate one basic activity, say,neutron diffraction. As the interest of the group enlarges and resources become available,additional groups of similar sizes should be formed following the same pattern.

2.3. INFRASTRUCTURAL FACILITIES

The research group, as mentioned above, can function properly only if its scientificefforts receive adequate additional infrastructural support. Essential requirements includeelectronics maintenance facilities, mechanical workshop, reliable electricity supply, an up-to-date library with recent journals, good computing support and radiation measuring and safetyfacilities. These facilities are normally needed to serve other scientific programmes andgroups as well. Specific requirements for various usages must be catered for at the planningstage and during subsequent improvements.

2.4. ESSENTIAL INVESTMENT NEEDS

Availability of experimental facilities is necessary for initiating any programme ofneutron beam research. At least one operating neutron beam instrument should, therefore,be made available to initiate a research programme, immediately after installation andcommissioning of the research reactor. A powder diffractometer would be a good startingpoint for this purpose. If financial resources do not permit procurement of a complete unit,efforts should be made to assemble one, by fabricating some parts and acquiring componentslike monochromators, detectors, parts of rotating devices, and control and data acquisitionsystems. This might be feasible over a period of 1-2 years. Additional assured funds shouldbe allocated to cover running costs — samples, sample holders, consumables, etc. — and formaintenance. If electrical power supply suffers frequent breakdowns, a guaranteed powersupply system must also be acquired.

The R&D activities may be expanded gradually by adding more instruments and staffover a period of several years. For optimum utilization of resources, the selection of otherinstruments should match the available neutron flux level of the reactor.

2.5. REVERSE MOBILITY OF SCIENTISTS

Training of scientists from developing research centers and association with researchlaboratories in industrialized countries play an important role for initiating and promotingR&D programmes. A useful step would be to encourage scientists from active centres in thedeveloped countries to work for suitable periods of time in smaller centres to give them aboost and provide continuity and acceleration in the pace of work. They could help to design,build, and install new instruments, and also imrove utilisation of available resources or takepart in cooperative research. They will be able to transfer their expertise more effectivelybecause of their familiarity with local environments.

2.6. INTERDISCIPLINARY COLLABORATION

Neutron beam research is interdisciplinary. It is of great importance to establish a goodcollaboration between various disciplines such as physics, chemistry, metallurgy and biologyat the center. This will help to identify a research programme of mutual interest and enhancethe awareness of interesting problems that can be solved by neutron techniques. Such acooperation will cause a more realistic working programme on one side and help promotethe use of neutron techniques.

3. FUTURE RESEARCH PROGRAMMES

One of the important questions that arise in planning a useful and viable researchprogramme in any center, especially at the initial nucleating phase, is related to the choiceof materials and scientific problems for neutron beam research.

The choice of a given research area in neutron beam work can be dictated by a numberof factors including: (a) local or national interests in specific materials; (b) the balancebetween applied and basic research interests and motivations; (c) industrial developmentinterests and (d) the interests and expertise of potential collaborators. This last aspect shouldreceive major attention — most highly successful research programmes will require morethan neutron investigations alone. The availability of local collaborators with expertise inmagnetometry, X ray diffraction, optical studies, study of macroscopic properties likethermal or transport properties, Moessbauer spectroscopy etc. and, above all, specifictheoretical knowledge, should have an important bearing in the selection of researchproblems. From the neutron scattering view point, the type(s) of instruments available forscattering studies (or that can be built at a given reactor) will limit the types of problems thatcan be tackled, as will the availability of high quality samples. Some specific commentsfollow on given experimental areas. The summary table below deals with types of condensedmatter systems that can be investigated by neutrons using specific techniques. The analysisis focused on only a limited subset of the possible systems that can be investigated withneutrons and is in no way assumed to be comprehensive.

The scientific information obtainable from various materials using various techniquesis outlined in the following sections.

TABLE I. MATERIAL SYSTEMS AND APPLICABLE TECHNIQUES

Materials

Magnetic compounds,polycrystallinesingle crystalPartially disordered magneticalloysMagnetic Fi'lms and supérlàtticësCeramic iïig'H Tc superconductorsOquias arid amorphous systemsGels, polymeric materialsGeological 'materials'Petrögenic materialsPolymer filmsZeolites, fast iö'fi conductorsTexture arid residual stressBiological systems

TechniquesDiffrac-

tion

•y

.............x„..........X

............x.............— — •••• — au — """""

...______x.............

XX.............x.............

SANS Reflec-tometry

X

X

'"""x""'"J

""•"x""'"'--T"--....„T........

..........x„........

__x..........

InelasticScattering

X..............................

.............x,._.........

Depolariza-tion

X

——y-"""

3.1. MAGNETIC COMPOUNDS

Much interest has been generated in recent years in the study of magnetic propertiesof intermetallic compounds, particularly those with rare earth elements, because of theirapplication in areas benefiting from a high magnetic energy product (e.g. Nd2Fe14B1; R2Fe17,etc.), as well as the unusual magnetic exchange and anisotropy properties of thesecompounds. Thus this class of materials can be studied from either an applied (industrial) orbasic research motivation. A strong team of collaborators is essential for success, includingpersons who: (a) make materials in either polycrystalline or single crystal form or both; (b)provide supplementary characterization information such as magnetization and Moessbauerdata, and (c) theoretical expertise, particularly if work on exchange and anisotropy propertiesis to be attempted.

The study of the basic atomic and magnetic structure of these materials inpolycrystalline form by neutrons requires only powder diffraction techniques to do successfulresearch, with the proviso that the analysis (say, by Rietveld analysis) must include magneticscattering.

If single crystal samples can be obtained, neutron scattering work can be expanded toinelastic scattering studies of the dispersive modes using a triple axis spectrometer todetermine exchange constants. Crystal field levels can, in principle, be studied onpolycrystalline samples for which the triple axis or time of flight method are appropriate.

3.2. MAGNETICALLY DISORDERED SYSTEMS

Study of the effect of disorder on magnetic interactions in materials is of considerablefundamental interest. However, it may not have direct practical applications at the presentstage of our knowledge. The disorder may originate from incomplete or random occupancy

10

of magnetic sites in a crystalline system, or from true topographic disorder such as is foundin an amorphous metallic glassy material. The disorder often leads to competingferromagnetic and antiferromagnetic exchange phenomena (spin frustration, etc.) and mayhave the effect of destroying long-range magnetic order, resulting in the formation ofmagnetic micro-domains. Magnetic phenomena in disordered systems are thus frequently ina size regime intermediate between nearest neighbor atomic dimensions and infinite rangecorrelations, and can be studied by small angle neutron scattering (SANS) for amorphoussystems, or by wide angle powder diffraction methods in the case of poly crystalline systems.The types of information that one obtains include spin correlation lengths, details of spinfluctuations above and below magnetic transition temperatures, etc.

In the class of disordered magnetic systems are many metallic glassy systems formedby rapid solidification (e.g. Fe^ Zr10, Crx Fe^,), ferrofluids (suspensions of finesuperparamagnetic particles in a liquid matrix) and magnetic nanostructures (e.g. Fe particlesembedded in an immiscible metal host or ceramic).

The question of the existence of long range order in alloys versus a short-range or spinglass order can also be studied by neutron depolarization. While this is principally aqualitative technique, it does provide valuable information on the magnetic state of the systemespecially when studied in conjunction with other parameters such as the magnetization andMoessbauer spectra. More quantitative results can be obtained from diffuse elastic neutronscattering.

3.3. MAGNETIC THIN FILMS AND SUPERLATTICES

This is one of the areas of magnetism currently of considerable interest. The ability toproduce thin films of arbitrary thickness, and to do atomic scale engineering of layeredstructures by molecular beam epitaxy (MBE) and sputtering, open up almost limitlesspossibilities to study exchange interactions, interface effects, and surface boundaryconditions.

The materials can be studied by a number of neutron techniques including wide anglediffraction, SANS (if disorder is present in a specific system), and by inelastic scattering ifsufficient material is available. With extremely high quality materials the technique ofneutron reflectometry can provide a wealth of information about variation of the magneticmoment near an interface, details of the interface roughening, and the range of the magneticinteractions.

The greatest limitation to work in this field is the very strict quality requirements themeasurements place on film materials. Thin films must be of good crystal quality, free ofimpurities, and uniform, to be of real use in neutron experiments. These are not trivialrequirements and often require very expensive preparation laboratories and labour intensivetechniques for both growth and proper characterization of the film materials. This is not anarea to be entered into if major resources (both equipment and personnel) are not availablefor sample preparation.

3.4. CERAMIC (HIGH Tc) SUPERCONDUCTORS

These materials lend themselves well to neutron powder diffraction techniques becauseof the critical dependence of superconducting properties on the location and occupancy ofoxygen sites which can be studied directly by neutron diffraction only. The materials are also

11

relatively easy to produce and can be studied from many different aspects — oxygendeficiencies, atomic site substitutions, bond lengths, etc.

3.5. TEXTURE AND RESIDUAL STRESS IN FABRICATED MATERIALS

Non-destructive evaluation of engineering materials by neutron diffraction is animportant field of investigation of materials reliability. Residual stresses induced in localizedregions during fabrication processes can lead to various deleterious effects like stresscorrosion or brittle fracture.

Neutrons, being highly penetrating particles, can examine the bulk stress distributionsby measuring strains in the materials. Careful wide angle diffraction measurements of theatomic d-spacing throughout small and controlled volumes of material near a weld or otherstrain-inducing deformation can provide the data to do a full three-dimensional analysis ofthe local stresses in a material and therefore an analysis of its potential for failure.

Powder diffraction techniques could be used to study texture in a variety of materials,like cast iron, rolled sheets, work-hardened stainless steels, etc. The anisotropy in physicalproperties is determined by the anisotropy of texture of the material under consideration, thatis the relative orientation of crystallites in the material. Conversely material properties canbe controlled by developing properly textured materials.

3.6. ZEOLITES

Zeolites are important catalysts in petrochemistry. Knowledge of how the absorbedmolecules are attached and interact within the zeolite channels are of interest from a varietyof points of view. Usually X rays provide the basic structural information. However,neutrons can "see" the hydrocarbon molecules in greater detail, and thus more usefulknowledge can be obtained from neutron powder diffraction.

3.7. PETROGENIC MATERIALS

Studies of petrogenic rocks are important both for the oil industry and for geology.Information about porosity and the surface roughness of the petrogenic rocks are of interestfor estimating the migration and formation process as well as to increase the efficiency of oilextraction and production. Much essential information for optimizing these processes can beobtained by measuring the fractal properties of the rocks using SANS techniques.

3.8. MINERALS, POLYMERS, AND AMORPHOUS SYSTEMS

Low and medium flux research reactors can be profitably used for characterization ofa variety of magnetic and non-magnetic materials. By characterization it meant thedetermination of relative composition and identification of constituents, for example, mineralsin a composite system by the use of neutron powder diffraction. The provision of this kindof support for mineralogical/geological analysis would be of interest to any reactor center.

For analysis of data one has to develop a data base of diffraction patterns of basicmineral constituents like MgO, CaO, MnO, Fe2O3, Fe3O4, SiO2, etc. The mineralconfiguration can, in principle, be obtained by a multiphase or a multi-component Rietveldanalysis.

12

Expanding the scope of these studies to fibrous materials like cotton, silk, or jute,would be of interest to examine the texture under various processing conditions to be ableto select the best processing parameters.

Yet another system of materials are the solid solutions or systems which showamorphous structure. Depending on the concentration/composition of solid solutions, thematerial can show more "crystal-like" behavior or more "liquid-like" behavior. Thesestructural aspects will govern various properties like conductivity, both thermal andelectrical.

4. METHODS AND INSTRUMENTATION

4.1. OVERVIEW

In view of the importance to acquire familiarity with the use of neutrons as a researchtool, at many of the existing low flux facilities, a well planned approach is needed. Thestrategy should be to start from relatively easy-to-implement methods, requiring little or nosophisticated equipment, and to proceed to more demanding ones as the competence of theteam grows.

The following list provides an outline of available techniques, their impact on trainingprogrammes, and ultimately their application in regular R&D programmes. Of course, it isnot mandatory to follow through all these steps to build up the required competence. Thisis presented only as a 'menu' to choose from, according to the particular situation andresources. For further details for each method, one has to consult detailed literature availablein various publications.

4.2. METHODS

4.2.1. Transmission-type measurementsTechniques: integrated transmission (cross-section) measurements

transmission imaging (radiography)neutron depolarization studies

Training benefits: use of neutrons in generalconcept of cross section, spin and magnetic moment of neutronsways of attenuating neutron beamsdetection methodsfilter techniques

Uses: transmission imaging (hydrogen distribution,macroscopic structural inhomogeneities, porosities)bulk magnetic propertiesquality control

4.2.2. Measurements on Bragg-peaksTechniques: double-crystal small-angle scattering

(inverted) Fourier correlation methodstatic Debye-Waller factors (DWF)texture determination

13

Training benefits: better understanding of neutron interaction with matter and scatteringtheoryhigher precision workmore advanced data analysis

Uses: large structural defects (of the order of 1 jam)microcracks in materialsfatigue and aging effectspolymer sciencenon-destructive testinglow hydrogen concentrationsstrain (integral measure through line width or spatially resolved throughline shift)lattice distortions due to defects (integral measure through DWF)effect of work process on materials (texture)

4.2.3. Determination of Bragg-intensities (diffraction)

Technique: two axis diffraction, with third axis to eliminate thermal diffusescattering (TDS) and other spurious effects

Training benefits: science related to background, beam collimation and advanced filteringcrystallography

Uses: crystal structure determinationeffects of compositional changes

4.2.4. Direct beam small-angle scattering

Technique: using narrow collimated beam based on low resolution wavelengthselection and position sensitive neutron detection

Training benefits: theory of scattering from small defects(isotopic) contrast variation data analysis

Uses: defects in materials precipitationslarge periodic structuresphase transformations

4.2.5. Diffuse-elastic scattering

Technique: extremely low-resolution "diffraction" between Bragg-peaks

Training benefits: elastic/inelastic scattering discriminationorder/disorder in materials

Uses: investigation of ordering in condensed matterstructure factors in liquids or quasi-liquids

4.2.6. Inelastic scattering using Be-filter technique

Technique: variation of incident energy with Be-filter in scattered beam totransmit only long-wavelength neutrons

14

Training benefits: inelastic scattering from materialsadvanced background suppression

Uses: incoherent inelastic scattering spectrum from hydrogenous systemsphonon density of states, especially heterogeneous compoundsphonon dispersion of high frequency branches

4.2.7. Inelastic scattering using time-of-flight or crystal monochromators techniques

Technique: Time-of-flight (TOP) technique or triple axis neutron spectrometry

Training benefits: measurement of inelastic spectralattice and molecular dynamicsmolecular, atomic, magnetic motion studiescrystal field effects

Uses: measurement of inelastic spectra from powdersphonon/magnon dispersion relation dataquasi-elastic scattering studies of reorientational motionscrystal field spectra

4.3. TECHNIQUES AND INSTRUMENTATION DETAILS

The methods listed above require an increasingly high degree of familiarity with boththe techniques of neutron scattering and the understanding of the underlying theoreticalprinciples as well as theoretical knowledge in condensed matter science.

They are also suitable for starting with simple equipment for data taking and analysisand proceed to more elaborate systems, building on a standard which should be introducedearly on.

Most of the results that can be obtained are of practical importance in the developmentof a technological culture and an industrial base.

In what follows, the individual methods will be discussed in more detail.

4.3.1. Integral transmission measurements

Measuring the neutron transmission through samples in the direct beam is the simplestform of a neutron experiment. Very little equipment is required: neutron absorbing material(cadmium, B4C-containing plastics) to limit the beam cross-section and divergence, a neutronmonitor in front of the sample and a detector together with the appropriate countingelectronics. In order to reduce the gamma-background, a filter should be installed in thebeam. Adequate shielding around the equipment to protect personnel should be a commonfeature of all experimental arrangements.

Apart from basic training in shaping neutron beams and counting neutrons (adjustingthe detector threshold, setting up a measuring arrangement, correcting for background) suchmeasurements can be used to determine the content of highly absorbing materials (Hg, Cd,etc.) or strong scatterers (especially hydrogen) with a relatively high degree of precision. Awell collimated arrangements can also be used to determine the degree of small-angle

15

scattering through its effect on the transmitted beam. Neutron measurements can also becarried out in a monochromated beam behind a crystal monochromator ("single-axisspectrometer") with the benefit of much reduced background and more quantitative resultssince the energy of the neutrons is known.

While such measurements will in general not justify the operation of the reactor in theirown right, the technique is very valuable for initial training and can well be developed intoa useful tool to determine the composition of specimens whose components are knownqualitatively but not quantitatively, if they weaken the neutron flux differently.

4.3.2. Neutron transmission imaging (radiography)

Transmission imaging, frequently also called "radiography", with thermal neutronbeams is one of the most straightforward applications of neutrons in materials research andtesting. Since neutron flux densities of the order of about 10s neutrons/cm2-s are sufficientto take radiographs in about 10 minutes, it can be done even at the smallest existing reactors.All one needs to perform such an experiment is a conical collimator (typical collimationangles for medium resolution work: 40') which provides for a uniform intensity distributionof neutrons emerging from a "point-like" source over the sample cross- sectional area(ranging in practice from just a few cm2 to about half a m2) and a film detector covered bya thin In foil. In the "direct" method the latter is necessary to produce conversion electronswhich in turn lead to an exposure of the photographic emulsion of the film. In the "transfermethod", which should be chosen in the presence of a large gamma background, oneactivates the In (or Dy) foil and subsequently transfers the activity image to the film, whichthen records the local variation of the emitted beta and gamma radiation. In high resolutionradiography systems with adequately well collimated beams (around 10') spatial resolutionsof the order of lOjam can easily be achieved if, in addition, a microdensitometric scanningprocedure is applied to evaluate the radiographs more quantitatively.

There is an almost unlimited range of possible applications of this technique. At thehigh resolution limit it has been used, for instance, for the investigation of 3H and 3Hebubbles in tritium-containing metals and the embrittlement problems associated with theirformation [1]. Radiographs of large objects with low spatial resolution, on the other hand,are quite useful on an almost industrial scale. A good example of this kind is theinvestigation of the humidity transport and humidity distribution in building materials, whichis hard to obtain with other methods. Even biological systems can be studied with neutronradiography. For instance, one could study the transport of water in plants by adding just afew percent of heavy water when watering the plant. There, one simply exploits the drasticdifference of the total scattering cross-section of hydrogen and deuterium. This (large)isotopic contrast is not the only important advantage of neutron radiography as compared toradiography using X rays. Almost equally important is the ability of neutrons to penetratethick objects without severe absorption.

In metallurgy, neutron radiography is a valuable non-destructive method. It can beapplied as a high resolution neutron radiography technique capable of investigating themicrostructure and composition of a metallurgical specimen both in a qualitative and aquantitative manner. This technique is called microneutronography. It can be used either forviewing the internal microstructures in the sample or possibly be applied in a quantitativemanner for the determination of the internal composition of a specimen by means ofmicrodensitometric recordings. An image is obtained by placing a thin specimen (100 /^m toa few 1000 /«m) in close contact with a suitable neutron imaging detector and exposing it to

16

a collimated beam of neutrons. The microneutronograph is then obtained by enlarging theprimary image by means of a light microscope. The complementary method tomicroneutronography for the determination of boron distribution is neutron-inducedautoradiography. In cases where a specimen contains elements like B, Li or U, which byitself emits charged particles upon neutron absorption, the presence of these elements can berevealed without any intermediate converter screen directly by detection of reaction particleswith a suitable detector.

4.3.3. Diffuse elastic neutron scattering (DENS)

Diffuse elastic neutron scattering (DENS) spectrometers are in principle very poorresolution diffractometers with a wide range of application. These rank as low priorityinstruments in the sequence of introduction of the various experimental methods because:

• the cross-sections involved are much lower than for Bragg reflections, so the techniqueis more flux-limited,

• to compensate for low cross-sections it is necessary to use many detectors at differentangles which makes such an instrument more expensive,

• thermal diffuse (phonon-inelastic) scattering and diffuse elastic scattering being of thesame order of magnitude in many specimens, a coarse time-of-flight selection or energyanalysis by other means is convenient but adds considerably to the complexity of theinstrument.

DENS investigates the scattered intensity between the Bragg peaks which results fromthe presence of different kinds of constituents in a crystal. In a binary system the intensityis determined by the difference in scattering amplitude of the (two) constituents and theirrelative concentrations:

I ~ C A . ( l - C A ) . ( b A - b B ) 2

This intensity is modulated as a function of momentum transfer Q in characteristic waysif the distribution of the atoms on the lattice sites deviates from random, i.e. differenttemperature treatments of materials can give information on the effect of this treatment onthe microstructural constitution of the material, or it can be used to analyze its state, if thetreatment is not known. If the method is used for non-destructive testing, it may allow tocharacterize specimens either on the basis of experience or on the basis of a theoreticalinterpretation of the scattering pattern.

DENS spectrometers, when using a sufficiently short wavelength, are also suitable forthe investigation of liquid structure factors at low Q. They will, however, not cover a wideenough Q-range to avoid truncation effects in a Fourier inversion to get radial distributionfunction.

The equipment needed is, apart from the usual beam shaping, shielding and monitoring,mainly a coarse monochromator (mechanical velocity selector or double crystalmonochromator), a low resolution beam chopper (for time-of-flight) or large mosaicanalyzer crystals in front of the detectors.

If crystal analyzers are used the range of incident wavelength will be limited unlesssubstantial mechanical effort is provided to equip the analyzers with variable energyadjustment possibilities. If time-of-flight is used, a low resolution multichannel time-of-flight

17

analysis should be performed, requiring a multidetector time-of-flight equipment.Alternatively simple gating of the detectors to the time intervals when the elastically scatteredneutrons arrive, may be a cheaper way but this limits the possibility of correcting forinelastic contributions to the measured intensities.

While being of somewhat higher complexity, DENS is a very useful tool in materialsscience and technology.

4.3.4. Neutron depolarization

Although the production of polarized neutrons is inevitably associated with a loss inintensity of at least about one order of magnitude, one can use the polarized neutrons forsome special purposes even at low flux reactors. In particular, neutron depolarization, whichis conceptually a simple transmission technique, is quite a versatile and sensitive tool to studymagnetic structures on mesoscopic scale.

In its simplest way of realization, one places the specimen under investigation betweena polarizer and an analyzer (which, for thermal neutrons of wavelength of less than 2.5 A°,are preferably magnetically saturated Heusler crystals; for longer wavelengths polarizingsupermirrors are the appropriate choice) and measures the variation of the degree of neutronpolarization on transmission through the sample by observing the change of the intensitybehind the analyzer. From a very simple algorithm, which dates back to 1941 [2], one canthen derive the mean size of the magnetic domains within the sample. In many respects thismethod resembles the investigation of transparent objects by placing them between polarizingoptical filters.

With a somewhat more sophisticated equipment, which essentially involves a 3-dimensional analysis of the neutron polarization [3], one can extract more information aboutthe domain structure, as e.g. the mean square direction cosines of the domain magnetizations,the mean macroscopic magnetization of the sample and the presence of various types ofcorrelations between adjacent domains. The main drawback of this technique, which hasprobably prevented it from being widely used, is the lack of simple, generally applicableevaluation algorithms to extract the domain structure information from the measured data ofsuch a 3-dimensional neutron depolarization experiment. Until now one needs supplementaryinformation from other techniques (magnetization measurements, neutron small anglescattering, electron microscopy etc.) to correctly interpret the experimental results in aquantitative way.

Its big advantage is its very high sensitivity to the presence of magnetic long rangeorder in the sample, which scales up exponentially with the square of the wavelength, and— since it is a transmission technique — its applicability at even very low neutron fluxes.Even at a reactor of typically 100-200 kW it allows one to study the domain structure ofmagnetic materials on a "real time" scale, that means it is possible to detect magneticrelaxation effects of the domain structure after periodic excitation by an external magneticfield with a time scale of several /us up to times which are only limited by the availableoperation periods of the reactor itself. For relaxation time-constants larger than about 1minute, it is, of course, even not necessary to vary the sample environment repeatedly. Thehigh sensitivity to magnetic inhomogeneities makes the neutron depolarization method alsovery well suited for the investigation of magnetic phase transitions.

18

4.3.5. Small-angle neutron scattering

Small-angle neutron scattering (SANS) is an important tool to reveal the microstructureand macroscopic properties of materials. It occurs whenever the inhomogeneities of10-1000 A° are present in any condensed substance- for example in the presence of pores orprecipitates in a matrix, macromolecules in solution, grain boundaries, submicron-cracks etc.These are important in homogeneities to characterize the microstructure of most materials:metallic, polymeric, composite, biological and geologic. The SANS technique providesinformation on the average size and number of inhomogeneities and also on their shape.Systematic SANS studies can also be carried out to optimize the processing of new materials.

It is fortunate that this technique, with such a wide range of applications, can beimplemented at a modest flux reactor. This is possible because:

• only a roughly monochromatic beam of neutrons is required, which means that arelatively wide slice of the neutron spectrum is actually used in the measurement;

• area position sensitive detectors are commercially available which can collect thescattering (at typically 4000 individual detector channels) simultaneously which isequivalent to increase the reactor flux by a few orders of magnitude; and

• simple multichannel collimators can be utilized in the incident flight path to makeeffective use of the full source area maintaining the small required divergence of theincident beam. For instance, a channel collimator with 4x4 channels is equivalent tohaving 16 SANS machines in parallel carrying out the same measurementsimultaneously, which means a gain in neutron flux of more than one order ofmagnitude. A further gain of 4 to 5 could be achieved through the use of a coldmoderator such as methane cooled down to liquid nitrogen temperature. This should beconsidered as a possible future development of a SANS facility.

As far as investment prices are concerned the most expensive components are the areaposition sensitive detector and the velocity selector for rough monochromatization. The lattercosts about US $10 000 whereas the detector costs about US $120 000 or US $220 000 foreffective areas of 25 X 25 cm2 or 64 x 64 cm2, respectively, both at a pixel width of 1 cm.The larger detector has a count rate which is 5 times higher than that of the smaller one forthe same measurement and the same neutron source. The remaining components can beconstructed locally if a sufficiently well equipped mechanical workshop is available.

(Note: The costs and sources in some cases from where some of the components canbe commercially procured only as examples. However these are not to be construed asrecommendations, the only suppliers or fixed authentic prices.)

A well performing SANS facility for research at a modest flux reactor would thusrequire the largest available area position-sensitive detector, the velocity selector and anoptimized design with a multichannel collimator which can be built locally. It should ideallybe installed at a tangential beam tube to avoid high levels of gamma-radiation and fastneutrons. In this case a scatterer has to be placed inside the beam tube which in a first stagecould be a water scatterer at ambient temperature. At a later date, this could be replaced bya cooled moderator such as methane. In any case the scatterer can also be constructedlocally.

A less expensive SANS using a single linear position sensitive detector for modestresearch could be built using a BeO filter acting as a coarse monochromator. The same setup

19

is upgradeable by replacing the linear position sensitive detector with a large 2-dimensionalarea detector as outlined above. Such a simple SANS has been in operation at DHRUVAreactor in India for research in a variety of surfactants, ferrofluids, etc.

4.3.6. Double-crystal SANS

To study inhomogeneitles, at the length scale of l /um (104 A°) such as precipitates andcracks, a simpler and less expensive SANS instrument can be built. It is based on the non-dispersive arrangement of two perfect crystals. The principle of the measurement is toanalyze the width of a perfect crystal reflection with the help of a second perfect crystal.Changes of the width introduced by a specimen in the beam path which causes small-anglescattering can thus be determined with high precision.

The main components of this instrument consist of two "channel-cut" perfect siliconcrystals, one of them fixed, acting as monochromator, while the other is placed on top of aprecise goniometer and rocked through the entire reflection, and a standard neutron detectorat a fixed position. The channel cut silicon crystals can be obtained with the assistance of theAtominstitut, Vienna at quite moderate costs, the goniometer can be such as the one used forstandard triangulation purposes and a neutron detector at a cost of about US $1000.

This is an interesting instrument to install at a modest flux reactor because neutron fluxcan be gained effectively by using a rather divergent beam incident upon the monochromatorset. Due to the fixed scattering geometry, SANS measurements can easily be carried out atdifferent temperatures, different surrounding atmospheres, with external magnetic and/orelectric fields, different pressures or tensile stresses, etc.

Thus the SANS technique applied either to the length scale of l /-im or to 0.1-0.001 /urn, offers an almost unlimited field of applications for materials testing and for bothapplied and basic research.

4.3.7. Two axis (powder) diffractometry

Two axis diffractometers (one axis at the monochromator for variable incidentwavelength and one axis at the sample for variable scattering angle) are the most widely usedinstruments for crystal structure determination and for investigation of a variety ofphenomena that affect the crystal structure. As a domestic tool, training device and forresearch, they are a valuable asset in every reactor neutron scattering outfit.

Neutrons from a (filtered) beam are monochromated with a crystal monochromator. Indesigning the diffractometer, take-off angles of 90° or more should be chosen to keep theoption for improved resolution open, as experience and demand increase. Also, thepossibility of inserting collimators in front of the monochromator and betweenmonochromator and sample must be foreseen, even if initially only a relatively coarsecollimation is used for intensity reasons.

In its simplest form, a well shielded detector is moved around the polycrystallinesample to scan the positions of Bragg reflections from the specimen. This should beautomated and requires mechanical gears of fairly high precision (at least 1/10 of a degree).It is of course an essential advantage and helps to overcome flux limitations in the reactor,if measurement is possible at many scattering angles simultaneously. This can be done byusing an array of many detectors or by going to position sensitive detectors. If many

20

detectors are used, corrections for sensitivity may be necessary but can be performed withthe help of standard incoherent scatterers. Position sensitive detectors are available as linearones, which are relatively cheap but count-rate-limited and require some corrections forscattering angle and resolution, or as curved ones which are rather expensive.

Although two axis diffractometers are comparatively simple to build and to operate,they have a very wide range of applications. It is possible to start with a rather simple setupof moderate resolution for testing and training purposes and improve the performance of theinstrument by exchanging parts of the spectrometers.

Air cushion technology is not necessary to move the detector arm but is veryadvantageous to increase the instrument's flexibility (e.g. by varying the sample-detectordistance, improving the precision of positioning, etc.). It also allows to add a third axis tothe instrument. On a low flux reactor the main reason for doing so would be to reducesubstantially parasitic contributions to the scattering by adjusting the analyzer to the elasticscattering from the sample. In favorable cases such an instrument may be used to measurealso inelastic scattering.

4.3.8. Fourier time-of-flight (TOF) neutron diffraction

Time-of-flight neutron spectroscopy of white chopped neutron beams is a standardmethod for high resolution powder diffractometry. However, with conventional single-slitneutron choppers it suffers from a bad neutron utilization at steady neutron sources becauseof the low chopper duty cycle.

With the so-called "Fourier-TOF" technique, introduced by Colwell et al. in 1969 [4],the situation is quite different. Here one does not measure the flight time of individualneutrons but modulates instead the incident neutron beam periodically in time with a seriesof discrete frequencies by means of a multislit disk chopper in combination with a similarlyequally designed stator and records by means of a phase sensitive detection system therelative decrease of the modulation depth and the phase shift of the observed intensityoscillations with respect to those produced by the chopper. By doing so one effectivelymeasures for each of these frequencies the Fourier amplitudes and phases of the time-of-flightspectrum, which can then easily be calculated by proper Fourier inversion.

The essential advantage of this technique is the high duty cycle of the Fourier chopper,which is 0.25 due to the 50% opening time and the 50% area losses of the rotor-statorcombination. Usually this technique was intended for inelastic scattering work, but it finallyturned out that with such a correlation procedure the absolute statistical error of all calculatedpoints of the TOF spectrum is the same and hence only those parts of the time-of-flightspectrum which lie above the mean intensity can be determined more accurately within agiven measuring time than with conventional TOF. Clearly this means that the elasticscattering dominates the much smaller inelastic peaks which consequently suffer from largerelative statistical errors. For neutron diffraction, however, where one is interested in theelastic peaks only the Fourier method may be of extreme advantage and is surely worthy tobe applied more often than it is up to now. The experimental requirements of this techniqueare very moderate and can be implemented practically at every laboratory irrespective of thelevel of sophistication of other techniques already in use.

In a modified version of this technique, the so-called "Inverted Fourier-TOF Method"invented by Hiismaki et al. [5] in Finland one can even achieve an on-line measurement of

21

the time-of-flight spectrum without the need of explicit Fourier-inversion of the data. Theadditional hardware requirements of this special technique are negligible. Since one mightwant to preserve also the possibility to measure the Fourier components of the spectrum, onecould easily design the spectrometer as to operate either in the "normal" or in the "inverted"mode, virtually without any significant extra costs to provide for this flexibility.

4.3.9. Inelastic scattering of neutrons

(a) T-O-F Method using a rotating crystal monochromator

Through neutron inelastic scattering experiments one has gained detailed informationon phonon spectra and dispersion curves, phonon lifetimes, thermal diffusion of atoms etc.,especially in the case when a sample contains atoms with a high incoherent or coherentscattering cross-section for neutrons.

When hydrogen atoms are present in the molecule, neutrons are scattered mainly bythese atoms, providing relatively easy interpretation. The magnitude of the energy transferinvolved in these phenomena is about 0.1 to 100 meV, which corresponds to interaction timesof 10'11 to 10~13 s. (Ultrasonic and dielectric relaxation methods give information on a timescale of 10"7 s and in nuclear magnetic resonance (NMR) measurements it is possible to reacha time scale of about 10"11 s. A time scale of about 10~12 s can be reached with infrared andRaman scattering but as the wavelength of the electromagnetic radiation is large, themomentum transfer is always almost zero).

The incoherent inelastic scattering spectra can be measured using cold neutron (coldneutrons have an energy lower than 5 meV) energy gain processes by a time-of-flight (TOP)method. As an example, at the 250 kW research reactor TRIGA in Ljubljana (Slovenia) witha maximum thermal neutron flux of 10I3n cm"2 s"1 a pulsed beam of almost monoenergeticneutrons obtained by means of a rotating lead monocrystal (AQ = 4.06 A°) is scattered by asample into a bank of 3He neutron detectors arranged at 4 different scattering angles to theincident beam direction at distance of 2-3 m from the sample. In the beam port a berylliumfilter at liquid nitrogen temperature is placed (10 cm x 8 cm x 30 cm). The beryllium filtertransmitting only the part of the neutron spectrum below the Bragg cut-off (5.2 meV) helpsto reduce the background and provide a coarsely-monochromated neutron beam. The energyspectrum of the scattered neutrons is analyzed by measuring the time-of-flight of the scatteredneutrons from the rotating crystal to the detector by means of a multichannel analyzer. Themomentum transfer in the collision depends on the scattering angle. Because we are facedwith the problem that the flux of neutrons in the low energy tail of the thermal spectrum isquite small in a low to medium flux reactor, a shift of the Maxwellian distribution to longerwavelength by means of a low cost liquid nitrogen cooled moderator is of great advantage.For neutrons of 4 A° wavelength an increase of the flux by a factor of 4 is achieved incomparison with an equivalent water moderator if liquid methane 110 K is used as moderatorsubstance. Such a liquid methane moderator can be a useful source of cold neutrons onresearch reactors with a power of up to 1 MW [6].

Together with the nitrogen cooled methane moderator the rotating crystal time-of-flightspectrometer represents a very valuable experimental tool for research in solid state physics,chemistry or biology. To build such an instrument is not an extensive task because only amonocrystal, a bank of 3He detectors and a multichannel analyzer are required besides somevacuum pumps and proper shielding materials.

22

(b) Inelastic scattering using a triple axis neutron spectrometerAnalysis of energy of monochromatic neutrons scattered by a sample can be achieved

using an analysing crystal and a detector. The instrument used is a Triple Axis NeutronSpectrometer, which is the workhorse at several medium and high flux reactors. In a TripleAxis Neutron Spectrometer, a monochromator is mounted on axis I, sample at axis II, andan analyser at axis III. This is followed by a detector and a monitor in the monochromaticbeam. The data acquisition system can be a PC-based motor control system and a simplecounting system. This allows to control various angular devices, independently of theinstrument. Since no T-O-F technique is used, a multi-channel system or a device like arotating chopper are not required. The TAS can be built as a natural upgrade of a two-axisdiffractometer and can be used for a variety of purposes:• as an 'elastic' diffractometer when the analyser wavelength is the same as 'incident

wavelength',• to study phonon or magnon density of states from powders,• to study crystal field excitations from powders,• to study dispersion relation of phonons or magnons using single crystal samples and• to study small angle scattering at any incident wavelength by positioning

the sample in between two 'channel-cut' single crystals one situated at the II axis andanother at the III axis.

The instrument has the unique advantage of being able to measure S(Q,w) the scatteredneutron intensity at any momentum transfer and any energy transfer as required. Hence, fordispersion relation measurements or for obtaining S(Q,w), say, for a liquid, one can have'constant-Q' scan or ' constant-w' scan or any general scan along any direction in (Q,o/)space.

By replacing the single crystal analyser by a long analysing crystal and the conventional10BF3 or 3He detectors by a linear position sensitive detector, one can improve the throughputmany fold. So also, one can design specific variations of the instrument for multi-analysispurposes or for converting the instrument for a 'filter-detector' spectrometer.

(c) Inelastic scattering using Be-filter techniqueA two-axis spectrometer (or a TAS suitably modified) can be used for inelastic

scattering studies if one imposes a filter between the sample and the detector. The filter couldbe a poly crystalline Be or BeO block (having cut offs at 5 meV or 3.5 meV respectively) andneutrons can be detected in the conventional mode after getting transmitted through the filteror a Be-BeO combination energy window (of 0.5 meV around 3.7 meV) if the detector isreplaced by an annular detector around the filter combination to accept the scattered neutronsfrom the filter combination.

Since the filters provide a coarse fixed energy analysis window, inelastic spectra fromsamples are measured by varying the incident energy of neutrons from the monochromator.So unlike a diffractometer, where the incident energy is held fixed and only an angulardistribution of scattered neutrons is measured by rotating the detector about the sample, herethe geometry of scattering angle is generally held fixed at 90° and the entire spectrometer ismoved around the monochromator to change incident energy.

The throughput of the instrument is better than that of a TAS and excitation energiesfrom various compounds can be easily measured. However one cannot select a specific pointin (Q,w) space.

23

4.4. DATA ACQUISITION AND CONTROL

Since the fields of electronics in general and small computers in particular are underextremely rapid development, on the one hand, and the need for data acquisition and controlcapabilities, on the other hand, may increase as the spectrometers develop at a given reactor,it is important to adopt a strategy which allows a modular architecture based on establishedstandards which can be expected to be around and supported also in the future. This, at thesame time, ensures compatibility between hardware and software on a broad scale. Importantadvantages of such a strategy are:• efficient use of the available technical staff,• interchangeability of modules between different instruments,• expandability on the basis of locally proven and established components.

4.4.1. Requirements for a modest structure

Use of a PC to control the experiment should be aimed at in all cases. For a system oflimited capabilities it may be sufficient to work with add-on boards on the motherboard.Many different variations of such add-on boards are available serving a large range of needs,such as motor control, temperature regulation, etc. Many of them come with goodstandardized software and can be programmed in BASIC. Interfacing of external electronicequipment can also be done via add-on boards which provide multiple TTL and/or analoginput/output (I/O) ports.

4.4.2. Requirements for multiple or more complex structures

As the need for control functions increases with increasing functionality of theinstrument, space for add-on boards in the computer will not be sufficient, nor will there beenough specialized boards available. Nevertheless, PC control of the spectrometers is stillpossible but external devices should be connected. They can control the fundamentalfunctions of the experiment in a dedicated way such as pulse counting, positioning ofcomponents, control of the sample environment such as temperature, pressure etc. Also dataacquisition from position sensitive detectors can be accomplished by such external modules.

The modular nature of such a system allows to decouple the various functionscompletely and provides for a large degree of flexibility in testing, maintenance andinterchange of compounds between different instruments.

4.4.3. Standard of communication

The modular nature of a control system relies crucially on the use of a well establishedand widely supported standard for the computer and the external devices. Standards arecommercially available, and widely used systems should be preferred. Such standards areRS232 or IEEE488, an industrial standard also called GPIB (General Purpose Interface Bus).We recommend to settle for these. GPIB has a high data transfer rate and allows multipleconnections on the same cable. The cost of commercial products with GPIB standard issometimes rather high, but, as long as the standard is respected, home-made solutions alsocan be incorporated. Such solutions may be available at other laboratories and informationinterchange should be encouraged, e.g. through brief reports in "Neutron News", etc.

With the electronic functions of the 'home-made' device realized on a printed circuitboard, it must be connected to a central processor unit (CPU) board by TTL input-output

24

ports. The CPU board can be external to the computer but it must be equipped with IEEE488and/or RS232 interfacing and its PROMs (programmable read-only-memories) must containthe resident command interpreter specially written for the respective application. In this waythe device can be tested and developed using any computer and is also portable to otherinstruments or laboratories.

5. CONCLUSIONS

When considering the utilization of a research reactor (independently of the originalmotivation for acquiring such a facility) one should realize that, in a developing country, itis often the largest single facility for scientific R&D. Its utilization is usually limited bysevere constraints related to financial, material and human resources. Therefore, mostoperators are satisfied if a reactor is operated with moderate success. Even in this situation,it can make a significant contribution to what may be called the 'scientific and technicalculture' at the national level.

Nuclear techniques have a wide range of multidisciplinary applications and can havesignificant economical impact in several areas. The production of radioisotopes for medicaland industrial purposes, environmental studies, assessment of impact of industrial activities,acquisition of know-how related to nuclear power, etc., are some of the common examples.It is therefore understandable why the number of research reactors is steadily increasing inthe developing countries. On the other hand efforts are needed to enhance research reactorutilization beyond the simply 'satisfactory' level in order to make better use of thecorresponding investment, human and material.

The report provides guidelines for initiating new programmes in neutron beam researchand for upgrading ongoing activities in this field. It will be useful for defining appropriateresearch programmes, and to make more cost effective use of limited financial and humanresources, the available equipment and the research reactor itself.

REFERENCES

[1] BUCHBERGER, T., RAUCH, H., SEIDL, E., Kerntechnik 53 (1989) 215.[2] HALPERN, O., HOLSTEIN, T., Phys. Rev. 59 (1941) 960.[3] RIETVELD, M.T.H., Z. Physik 259 (1973) 391.[4] COLWELL, J.F., LEHINAN, S.R., MILLER, P.H., WHITTEMORE, W.L., Nucl.

Instr. Meth. 77 (1969) 135.[5] Hoiismaki, P., Juntilla, J., Piirto, A., Nucl. Instr. Meth. 126 (1975) 435.[6] Dimic, V., Petkovsek, J., J. of Phys. E 4 (1971) 905.

Next page left blank 25

Annex

PAPERS PRESENTED AT THE ADVISORY GROUP MEETING

FUNDAMENTAL AND APPLIED NEUTRON RESEARCHAT THE 250 kW TRIGA REACTOR VIENNA

G. BADUREKInstitut für Kernphysik,Technische Universität,Vienna, Austria

Abstract

A review is given about the research activities at the 250 kW Vienna university low fluxreactor, which are adequate for implementation at other neutron souces of comparable size.The topics selected for presentation range from neutron radiography, materials irradiation,neutron small-angle scattering, neutron activation analysis, neutron depolarization in magneticmaterials to neutron diffraction experiments for structure investigations of condensed matter.

1. Introduction

As it is the aim of the present meeting to stimulate programs for more efficient useof low power nuclear research reactors with neutron flux densities of the order of 1013 cm2

at the center of the reactor core, we briefly describe the experimental facilities installed at the250 kW TRIGA reactor of the Austrian Universities in Vienna (Rg. 1) and present asummary of a great part of the current research activities performed with them. We believethat most of the techniques and experiments presented here are adequate for implementationat other reactors of similar or even higher power. Those technologies which require extremelyspezialized know-how not generally available at every research reactor institution will not betreated here or are just mentioned without any further details.

It is common knowledge that due to the relatively low neutron fluxes of such reactorsone of the most important applications of neutron scattering on condensed matter, namely thestudy of atomic and molecular dynamics of solids and liquids, a priori must remain out ofconsideration. However, this does not mean that it is not possible to develop new or toimprove existing techniques for such experiments at low power research reactors. In fact suchdeveloping work has always been a crucial point of our research efforts in the variety offields of applied and fundamental neutron physics we were engaged in. But for the sake ofconciseness here we will concentrate us just to a "grosso modo" description of theinstruments and methods currently available at our institution. The order in which theindividual topics will be treated is roughly correlated with both their growing experimentaldifficulty and their increasingly less general applicability.

2. Neutron Radiography

Neutron radiography is .one of the most typical applications of small reactors becauseof several reasons. Firstly, it allows to perform useful experiments even with moderate

29

TRIGA-REAKTOKWIEN

HOY na

Fig. 1: Sketch of the experimental facilities installed at the Vienna University reactor.

neutron fluxes. Secondly, due to its conceptual simplicity it can be realized quickly withoutthe need of extremely specialized and trained personnel. And last but not least, it can beconsidered as an immediate and almost always profitable step towards material orientedresearch.

Two radiographie facilities exist on the TRIGA Vienna, one is installed at the thermalcolumn and mainly used for high resolution investigations of small samples not larger thanabout 9 cm in diameter. The other serves for inspection of larger objects (o^ = 35 cm) onan industrial scale (Fig.2). Conical collimators are installed at both facilities to produceuniform collimations and intensity profiles. The gamma background of the reactor is reducedby means of bismuth filters (polycrystalline, 4 cm at system 1; single-crystalline, 14 cm atsystem 2). The walls of the collimators are coated with materials containing boron andlithium to suppress (n,y)-radiation. The geometrical beam collimation, the neutron fluxdensity at the sample position and the gamma dose rate are as follows:

• system 1: 26', 1.5 x 105 cm'2 s"1 , 0.02 Gy h'1

• system 2: 75', 2.6 x 105 cm"2 s"1 , 0.1 Gy h'1'

30

«weIm

conical Si-filtercoUimator

Fig. 2: Cross-sectional view of the two radiagraphy facilities.

The exposure times are of the order of 10-20 minutes. The darkening of the radiographsobtained by the usual gadolinium converter film technique due to the gamma background isalways less than 20%.

As a typical result obtained with system 1 Fig.3 shows radiograpic images of a set of electronicmultilayer capacitors. These capacitors were indendedfor the German "Meteosat" weather-satellite projectand we had to be investigated for eventual structuraldefects. In some of these capacitors air bubbleinclusion is clearly seen on the radiographs, whichwith high degree of probably would have led to amalfunction of the satellite electronics at the orbitTwo sets each consisting of 250 capacitors werechecked, almost 40% of the first set were found to

Fig. 3: Radiogaphs of electronic have some obvious defects. The second set was foundcapacitors revealing some structural to be 100% free of defects. One of the more recentdefects of the cast epoxy resin. pfojects was ^ investigation of 3H and 3He bubbles

in various metals containing tritium, which isinteresting in particular with respect to the embrittlement of structure materials of futurenuclear fusion reactors [1]. There the radiographie films subsequently were evaluated morequantitatively by microdensitometric scanning. The achieved spatial resolution was about 16pm. On the other hand one of the typical applications of the large scale radiography facilitywas the study of the humidity transport in bricks and other building materials.

3. Materials Irradiation

It is well known that the physical properties of many materials change upon neutronirradiation. In general this is of disadvantage but there are also some examples of profitable

31

use of this effect, in particular for semiconductors and superconducing materials. Controlledneutron irradiation of semiconductors can be used to modify their doping profiles, whereasirradiated superconductors exhibit inreased critical current densities compared to their initialstate. In many cases this increase of the critical current, which is still one weak point ofpresently available ceramic HTC superconductors, reaches more than one order of magnitude.There neutron irradiation at different temperatures produces different types of pinningcenters, which cause in turn a variety of changes of the superconducting properties that canbe either stable or exhibit reiaxational behaviour. In any case such investigations can help inour basic understanding of ceramic superconductors, but are also of importance for increasingtheir performance. The low temperature group of our institute has spezialized to this kind ofirradiation investigations of superconductors, which are also of relevance for the developmentof superconductors sustaining the extremely high neutron radiation dose rates in fusionreactors. An example of the increase of the critical current of a HTC superconductor byneutron irradition is given in Fig. 4, which also shows for comparison the completeindependence of the critical current of a sputtered NbN thin-film sample on the neutronfluence [2].

10

10'

Sample Cooooo in-., cold tranofar»»*»» irr., after wmrm— up

1O 15B (T)

Fig. 4: (/) Increase of the critical current of a HTC sample upon neutron irradiation.(r) Irradiation independence of Jc of a NbN film.

4. Neutron Small-Angle Scattering

Standard neutron small-angle scattering (SANS) facilities are large instruments,which often are installed at the end positions of cold neutron beam guides of high andmedium flux reactors, which usually are equipped with a Cold Source providing longwavelength neutrons. On the other hand, with the small-angle camera based on the non-dispersive arrangement of two perfect crystals, which is shown schematically in Fig. 5, onecan achieve with thermal neutrons an extremely high angular resolution of a few seconds ofarc, corresponding to momentum transfers of about 10"4 A"1 which is about two orders ofmagnitude better than that of standard SANS instruments [3]. Since the angular resolution ofsuch a small-angle camera is completely decoupled from the divergence of the incident beam

32

a rather divergent beam can be used, providing sufficient intensity at medium and even lowflux reactors. To suppress the disturbing side tails of the resolution function by multiplereflection we have equipped the camera with two "channel-cut" perfect silicon crystals asindicated in Fig. 5. The compactness of this land of small-angle camera allows to performreal-time experiments to investigate for instance ageing and relaxation effects in the samplesunder investigation. This technique allows to study the formation of any kind ofinhomogeneities, as e.g. preciptitates or cracks, with spatial dimensions of up to about 1 urn.Measurements at different temperatures, different surrounding atmospheres, with externalmagnetic and/or electric fields, different pressures or tensile stresses etc. can easily be

CHANNEL-CUTCRYSTALS

SAMPLE

DETECTOR I

DETECTOR 2

Fig. 5: Sketch of our perfect crystal SANS camera and comparison of its momentumresolution with that of two similar systems in Prague and Grenoble.

STANDARO RESOLUTIONLESS BACKGROUND

^CHANNEL(CUT

SMALL SAMPLESLESS RESOLUTION

.FANKUCHENCUT

LARGE SAMPLESHIGH RESOLUTION

CHANNEL CUTHOHOCHOOHATOaCRYSTAL

SAMPLE

BENT COMB-SHAPeOANM.YZEP CRYSTAL

Fig. 6: Possible modifications of the standardsmall-angle geometry for more efficient use of theavailable neutron intensity.

accomplished and thus offer an almostunlimited field of applications both forapplied materials testing and basicresearch. As indicated in Fig. 6 onecould modify the scattering geometryof this camera according to the specialrequirements of the sample system.And by replacing the standard channel-cut analyzer by a bent comb-shapedcrystal it should be possible incombination with a position sensitivedetector to measure the angulardistrubution of the transmitted beamsimultaneously without resorting to theusual rocking curve measurement [4].

33

5. Neutron Activation Analysis

COURTS

Besides of production of short-lived radionuclides partly for commercial purposes andpartly for other scientific institutions the main application of the Vienna reactor in the fieldof radiochemistry is the neutron activation analysis of a broad variety of sample systems.Applications reach from the detection of environmental pollution, the analysis of filter dusts

of fossile power plants, trace elementanalysis of geologic materials in searchof mineral deposits to - for instance -the authenticity check of antiqueobjects. An example of this latterapplication is shown in Fig. 7, where itcan be seen that in the falsification ofan antique copper coin in contrast to thegenuine one, dating back to the ancientRoman empire, the alloy contains nonoble metals. Two pneumatic transfersystems have been installed at ourreactor for the purpose of neutronactivation analysis. A slow one withtransfer times of several seconds and afast one with a transfer time of thesample from the reactor core to themeasurement position of only 120 ms,which allows for the detection ofradionuclides with short decay times ofjust a few milliseconds. Whereas the

""* '"" ""* slow standard transfer system is easy toFig. 7: Authencity check of an antique coin. In the impiement at any reactor the fast onefalsificate (below) the noble metals are missing. deseryes ^ development of special

electronic equipment to handle theextremely high gamma-count rates associated with such short-lived radionuclides [ 5 ].

coums

6. Neutron Depolarization

Its origin dating back in the early 'fourties neutron depolarization is a quite wellestablished technique to study the domain structure of magnetically ordered materials [6].Since it is essentially a transmission method it can be successfully applied even at relativelyweak neutron sources, although the production of polarized neutron beams inevitably impliessome loss of intensity. Without going deeply into details Fig. 8 shows the main componentsof the three-dimensional neutron depolarization setup which was installed at a tangentialbeam tube of the Vienna university reactor a couple of years ago and which since then was

34

used routinely for a variety of investigations of different classes of magnetic materials [71.By this facility it is possible to reveal the influence of temperature, applied magnetic fieldand/or tensile stress on the mean size and orientation of the ensemble of magnetic domainswithin the sample. Relaxation effects of the domain structure after sudden variation of theexternal field or the applied stress can be investigated on a timescale ranging from about 100ps up to several hours. Its function can be briefly described as follows:

MAGNETICSHIELDING GUIDE FIELD

\ DETECTORr- /

/ANALYZER

}START J.CH ADV

Fig. 8: Schematic sketch of the 3-dimensional neutron depolarization setup.

The incident beam is monochromatized (X = 1.5Â) and polarized by Bragg reflectionat a magnetically saturated Heusler crystal (P0 = I P0 I •• 0.95). On transmission through thesample, which is mounted within a magnetically shielding soft iron box and whosetemperature can be varied between 4.2 K and about 800 K by means of a either a cryostat ora quartz furnace, the polarization P0 of the incident beam changes to the final state Pc

according to the relation

where the so-called "depolarization matrix" D depends on the details of the magnetic domainstructure within the sample. Two spin rotation systems, which consist of specially designedcoil combinations described in detail in the literature [8], are attached to the shielding box.They serve to orient the incident polarization vector successively in any of the three spatialdirections and to project any of the three components of the final polarization onto thedirection of the analyzing crystal in front of the detector. In combination with a spinflipdevice which allows for a sudden reversal of the incident polarization state one is then ableto measure all 9 elements of the (3 x 3) depolarization matrix D. By means of mathematicalinversion algorithms it is then possible to derive the essential parameters of the domainstructure, namely the mean size of the domains, the mean square direction cosines of thedomain magnetizations, and also to obtain at least qualitative information about eventualcorrelations between neighbouring domains. Fig. 9 summarizes the action of a set ofhomogeneously magnetized domains on the transmitted neutron polarization vector.

35

From the many experiments we have performed with this neutron depolarization setupFig. 10 shows the result of an analysis of the influence of tensile stress on the domainstructure of amorphous ribbons as measured with neutron depolarization and the electro-optical Kerr-effect. Again neutrons offer the advantage to "see" inside of the samples, notjust their surface. One should notice the extreme sensitivity of the depolarization method, asfollows from a comparison of the depolarization matrix calculated from a theoretical model(Fig. 10, middle) with the experimentally measured one. Even minute variations of the modelparameters lead to a significant deviation of the theoretical from the experimental data atleast for one of the nine matrix elements.

B =IB|np tT"O;

neutrons transmitting homogeneous field

Jl

\ I ** \ 1

1 2 3 N-1 N

neutrons transmitting N domains

1 2 3 N

n

P. =1["o<ni.o,tj) P,

\P=

j=1,M

P. = J)

neutron beam transmitting bulk materialD =

Fig. 9: The principle of neutron depolarization on transmission through ferromagneticdomains.

Another application of the neutron depolarization method which is of interest for theindustrial development and production of modern permanent magnets, is the investigation ofthe magnetic anisotropy of hardmagnetic Nd-Fe-B specimens [9]. It turns out that valuableinformation can be extracted from the depolarization results which is hard or impossible togain by conventional magnetization techniques. Very recently we have started a joint researchproject with a group of Sao Paulo university to develop a new setup allowing forinvestigation of magnetic relaxation effects simultaneously on one and the same sample byneutron depolarization and magnetization measurements.

For completeness we mention that the a similar setup has been used to demonstratefor the first time that geometric phase shifts of quantum mechanical systems, known as theso-called "Berry phase" [10], can be observed even for incomplete evolutions of theenvironment of that particular system, in our case a spin-Vi particle. To realize thisexperiment instead of the sample a rotating magnetic field was established along the flight

36

path between the two spin turn devices of the depolarization facility [11]. But such a highlyspecialized experiment, belonging to fundamental quantum mechanical research, is surely notadequate as a "beginners" experiment on reactors in developing countries.

7. Neutron Diffraction

Neutron diffraction is a tremendouslybroad field of research performed at pJmost allscientific neutron sources around the world.Nevertheless, to do competitive work at arelatively low flux reactor one has to treatproblems where it is not so essential to have ahigher flux. One obvious situation of this kindrefers to the case of structural relaxation effectsin any kind of solid matter with time constantslarger than about several hours, so that in areal time experiment a statistically sufficientnumber of neutrons diffracted by the samplehits the detector. Clearly there would be noessential gain of information if the neutronintensity increases further. Another possibilityis to chose ambiental conditions at the sampleposition which can be established periodicallybut always for short times only. If, as it is thecase with our TRIGA reactor, the neutronsource can be pulsed synchronously with thisperiodic variation of the sample environment,so that the whole neutron flux is concentratedat short periodic time intervals and thuscorrespondingly increased during them, one canstudy the atomic or magnetic structure of thesample under the chosen external conditionsvery efficiently.

According to this considerations wehave developed a pulsed field-neutrondiffractometer, which allows to study magneticstructures with atomic resolution in magneticfields of up to 20 Tesla [12]. Fig. 11 shows a

schematic sketch of this diffractometer which is installed at one of the two radial beam tubesof our reactor. A graphite monochromator selects a wavelength of about 1.1 A out of thethermal reactor spectrum. The sample can be cooled to liquid helium temperatures by astainless steel cryostat of only 17 mm outer diameter. This is sufficiently small to fit into thecentral bore (<{> = 18 mm) of a split-coil magnet which is connected to a 8 mF/2500 V

0 100 300 300 400 WO 200 300 (00 V» ZOO 300 <00

'yy

loo zoo 300 un 100 300 300 U3O 100 MO 300 400

100 200 900 400 0 100 200 300 400 0

TENSILE STRESS [MPa]«0 200 300 (00

Fig. 10: Kerr-effect and neu t rondepolarization studies of the domainstructure of amorphous ribbons (see text).

37

DE TBEAM.SHUTTER

capacitor bank. Discharging of this capacitor bank via a high power thyristor circuit producesa field of up to 20 T for typical time intervals between 5 ms and 15 ms, depending on thechosen wiring of the discharging circuit. This fits not too bad to the minimum pulse width

of approximately 40 ms of the TRIGA reactorin its pulsed mode of operation, which allowsto increase the reactor power and hence theavailable neutron flux by a factor of about1200 at a maximum repetition rate of about 5pulses per hour. Fig. 12 shows that at leastwith single crystalline samples it is possible tomeasure Bragg peaks even with a single non-repititive "shot" of the reactor. The ideabehind this kind of experiment is to detect thefield-induced breaking up of the spin structureof antiferromagnetic materials via theassociated change of the structure factors.However, the pulsed mode of operation of thereactor causes a thermal and mechanical stressof the fuel elements and it is hence risky inparticular with old fuel rods. Therefore onehas to be very careful in the preparation ofsuch an experiment to keep the number ofnecessary pulses as small as possible. Becauseof this fact and since the experimentalrequirements are rather sophisticated this kindof research is not generally applicable even ifthe respective reactors are principially able tooperate in a pulsed mode. As indicated in Fig.11 the pulsed magnet equipment can be usedalso for depolarization studies. There the highfield is essential to saturate extremelyhardmagnetic materials like Nd-Fe-B. Theirsubsequent magnetic relaxation and domainformation can then conveniently be studied byobserving the behaviour of the transmittedneutron spins.

PULSED 401'COIL'POWER SUPPLY <^>

Fig. 11: The pulsed field neutrondiffractometer.

35. i31.1

ZX 2S.I.HL a.iiA 15.1II U.I

S.I

I.I

h (100)h.

'1 1I 1

i1 ,1 1

B.8 li.l 1S.I 28.1 ZS.81-3Z

art ».s

Fig. 12: Antiferromagnetic (100)-refiectionof a MnF2 single crystal measured in 1 ms.

8. Neutron Interferometry

In 1974 the first interferometer for matter waves with macroscopic spatial separationof the interfering beams has been successfully developed and installed at our reactor [13].Since that time it was used fer a series of quite spectacular fundamental quantum mechanical

38

experiments [14]. Although the overwhelming majority of this experiments had to beperformed on the high flux reactor of the Institute Laue-Langevin Grenoble simply becauseof intensity reasons, it was essential to conceive and to prepare them at our reactor as wellas to be able to test the functionality of the various components of the final setup. And notto forget, we and our involved students could accumulate sufficient know-how to finallyrealize the experiments at the high flux reactor.

Again, facilities like a neutron interferometer definitely represent one of the mostadvanced technologies in the field of neutron scattering and from that reasoning anunexperienced staff at a low flux neutron source should not at all start with such a difficulttopic. Because of its complexity we will also not discuss it further here. Even only anintroductory review would go far beyond the scope of the present report. But on the otherhand it is an outstanding example of the importance of small reactors for the developmentand realization of new experiments and techniques, by giving young scientists the opportunityto improve their experience and by giving them a chance to follow unconventional ideas andconcepts, which is often not possible at larger reactor facilities with their ultimate need ofcost effective exploitation of the available beam time.

9. Concluding Remarks

We have shown that even at very small research reactors with low neutron fluxes avariety of valuable possibilities exist both for applied and fundamental research. Besides thetopics we have presented her one should not forget two other important activities which areabsolutely necessary to operate any nuclear reactor, namely radiation protection and reactortechnology. For both of them qualified staff must be available, which in general can hold itsstandard only if it is not only involved in running the reactor but is also engaged in state-of-the-art research. Furthermore we have also not treated here the possibility to develop and totest new techniques and facilities at a small research reactor which are intended for later useat high flux neutron sources. Always during the last three decades this was a major field ofeffort at our institute, the most outstanding example being, of course, the successfuldevelopment of the first neutron interferometer. And, once again, we finally want to stressthe importance of small reactors for the education of young scientists and for the preparationof experiments at large high flux facilities.

References

[1] T. Buchberger, H. Rauch and E. Seidl; Kerntechnik 53 (1989) 215.[2] H.W. Weber, P. Gregshammer and R.T. Kampwirth; IEEE Trans, on Magnetics 25

(1989) 2080.[3] A. Miksovsky, H. Amenitsch, H. rauch and E. Seidl; Phys. stat. sol. (a) 130 (1992) 365.[4] L. Niel, H. Rauch and E.v Seidl; J. Appl. Cryst. 24 (1991) 562.

39

[5] G.P. Westphal; Nucl. Instr. Meth. B 10/11 (1985) 1047.[6] O. Halpern and T. Holstein; Phys. Rev. 59 (1941) 960; M. Th. Rekveldt; Z. Physik

259 (1973) 391.[7] G. Badurek, J. Hammer and H. Rauch; J. Magn. Magn. Mat. 14 ( 1979) 231; A. Veider,

G. Badurek, H. Weinfurter and K. Stierstadt; Phil. Mag. 58 (1988) 573; A. Veider, G.Badurek, R. Grössinger and H. Kronmüller, J. Magn. Magn. Mat. 60 (1986) 182.

[8] A.I. Okorokov, V.V. Runov and A.G. Gukasov; Nucl. Instr. Meth. 157 (1978) 767.[9] G. Badurek, R. Giersig, R. Grössinger, A. Veider and H. Weinfurter, J. de Physique

Colloq. 49 (1988) C8-1831.[10] M.V. Berry; Proc. Roy. Soc. London A 392 (1984) 451.[11] H. Weinfurter and G. Badurek; Phys. Rev. Lett. 64 (1990) 1318.[12] R. Grössinger, G. Badurek, Ch. Gigler and Ch. Schotzko; Physica B 155 (1989) 392.[13] H. Rauch, W. Treimer and U. Bonse; Phys. Lett. A 47 (1974) 369.[14] 'Matter Wave Interferometry' (G. Badurek, H. Rauch, A. Zeilinger, edts.), Physica B

151 (1988).

40

USE OF RESEARCH REACTORS FORSOLID STATE PHYSICS STUDIES ATBHABHA ATOMIC RESEARCH CENTRE, INDIA

K.R. RAOSolid State Physics Division,Bhabha Atomic Research Centre,Trombay, Bombay, India

Abstract

Three research reactors, — Apsara, Cirus and Dhruvahave been utilised for over three decades in India for carryingout a variety of Solid State Studies. In this paper, we reviewseveral aspects of these studies especially related to theefforts during the past ten years. Plans were drawn around 1975

to design the neutron beam facilities of the-then proposed DHRUVACv2xl0iJ* ns/cm^-sec) research reactor to cater to solid statephysics studies. Beginning 1982, comprehensive actions were takento design the neutron spectrometers, detectors, guides and hotand cold moderators and fabricate them locally. The reactor hasbeen operating at high power levels since 1988 and theindegeneously developed neutron instruments have providedexcellent facilities for solid state studies. Details of theseplans, developments and results will be presented.

1 - In troductionApsara, Cirus and Dhruva are three research reactors which

have catered to research and development programs at Trombay,particularly for solid state studies during the past fourdecades. Their characteristics are summarised in Table I .

Apsara provided very useful means of training scientists inthe initial phases' of neutron beam research programmes.

Currently, it is being used for isotope production, neutronradiography and neutron detector testing. There are proposals toconvert this reactor into multipurpose higher flux research

41

Table I_ :_ Research Reactors for Solid State Research at TrombavReactor Type Max.

PowerMax.Fluxin core

Criticalitydate

Apsara Light water 1 MWenriched U—metal fuel

Cirus D3O moderated 40 MWHaO coolednat. U—metalfuel

Dhruva D3O moderated 100MWDatO coolednat. U-metalfuel

"10ian/cm2-sec Aug.4,1956

~6xl0A3n/cma-sec July 10,1960

Aug. 8,1985

reactor that can operate upto 10MW and provide fluxes of nearlyFeasibility studies are under way.

Cirus reactor was the main research facility for nearly twodecades (1962-1985) for Solid State Physics. A large number andvarity of neutron spectrometers were built and operated at thisreactor to investigate a variety of Condensed Matter problemslike structure and hydrogen— bonds in hydrated crystals and aminoacids, magnetic structure of several mixed ferrates, heusleralloys, dilute magnetic systems to map out magnetic electrondensity distributions, structure of simple liquids like Zn, Ga,CCI* etc., dynamics of atomic motions in hexagonal crystals Mg,Zn & Be, and molecular motions in NH*— halides, liquids Ct-L», Cl^,

NH3 and ND3 etc . A variety of new concepts in neutron beaminstrumentation like window filter, multi-axis (analysis)spectrometer, temperature gradient based high resolution windowanalyser, data acquisition and spectrometer control systems likefully automated computer controlled diff ractometer,microprocessor based control systems were designed, built andcommissioned for improved utilisation of the reactor. Table IIlists the spectrometers that were operating at this reactoraround 1980 and the results of main investigations.

42

Table II ̂ Spectrometers in use at Cirus Reactor- Circa 19BB

Spectrometer Use related to Main investigationsPowder Diffracto-meter

Single CrystalDiffractometer

-Powder diffraction —Characterisation-unpolarised neutron —magnetic structure ofdiffraction ferrites included

canted spin structure

Crystalstructuredetermination

Polarised NeutronSpectrometer

Magnetic structure

Triple A>sisSpectrometerMulti—axis InelasticSpectrometerFilter DetectorSpectrometer

Rotating CrystalSpectrometer

Lattice dynamics &molecular motions

do

For inelasticspectral studiesfrom powders

Study ofreorientationalmotions in solids?< liquids

-Structure of hydratedcrystals to understandhydrogen bondinteraction,-structure of sixamino acids,-structure variationwith temperature eg.of LiKSCUMapping magnetic momentdistribution inferrites & dilutemagnetic alloys likeNi—Ru systems

-Phonon measurements inMg,Zn,Be,CaO & KNO3-elastic diffractionfrom ammonium ha1ides

Motion of watermolecules in hydratedcrystals & simplebiological molecules-Dynamics of liquidsa-U,CD*,NH3,ND3—Dynamics of NHV &CHa"*" in ammonium andmethyl salts

A large number of visiting scientists from countries likeIndonesia, Phillipines, Thailand, Korea, Egypt, Poland and Iraq

worked at this reactor and some of them stayed on for long enoughperiods for doctoral degrees based on work carried out at

Trombay. Indian scientists also took part in the Agency'sRegional Cooperation Agreement Programme for 5 years at Manilla,Phillipines to train scientists and initiate neutron scatteringprogrammes in many of the countries in the region. Severalspectrometers or neutron scattering related apparatus werefabricated at Trombay and supplied to the neighbouring countries.

43

Cirus reactor has continued to operate 'in its 33^«* year and

has been a very useful reactor for our many programmes. Currently

a small angle neutron spectrometer, a neutron interferometer, a

position-sensitive detector test facility and a neutronspectrometer for demonstration and training purposes have

replaced some of the existing spectrometers but otherspectrometers have continued to be used for research.

Dhruva reactor has been serving as an additional more

powerful source of neutrons since it went into full power

operation in 1988. As already stated, solid state physicists were

involved in the planning stage of this reactor unlike the case

with Apsara and Cirus. This helped us in overcoming and reducing

limitations of low neutron flux which would have been a bane

otherwise. We shall detail some of these aspects in the following

section.

2. Means for overcoming or reducing limitations of neutron f luxat Dhruva reactor

Discussions related to planning Dhruva reactor were taking

place in the early seventees, when several medium flux and high

flux European facilities had already gone into operation with

their very sophisticated neutron beam instrumentation that

included neutron guides, hot and cold moderators, guide halls,

multi-detectors etc. The flux of these reactors were in the range

of IB1** - 1B1!» ns/cm3-sec. Hence, it needed courage and tenacity

to embark on design and development of a medium flux neutron beam

research facility at the then proposed DHRUVA reactor whose

characteristics were largely determined by the use of natural

uranium metal fuel in a heavy water moderator, coolant and

reflector environment. This limited the maximum flux of the

reactor to be < 2>:10A* ns/cm2-sec and also it was to be a large

44

volume reactor like Cirus. Inspite of these limitations the solid

state physicists at Trombay ventured to plan their R&D facilities

with due consideration paid to installation of proper in—pile

beam features as well as out—of—pile spectrometer

characteristics. Some of the options examined and implemented at

Dhruva Reactor are given in Table III.

Table III In—pile beam options and ont—of—pi1e features

(i) In-pile Beam Features— Cut-aways in—pile shielding backed by high density

shielding— in—pile cavities for installing monochromators to use

pile shielding as part of monochromator shielding— installation of a through tubes containing a section of

DaO to reduce fast- neutron and gamma radiation in theneutron beam

— tangential beam tubes to reduce fast neutron component— provision to instal guides, cold source and hot source

(ii) Dut-of-Dile Features— large monochromator—drums backed with yokes to reduce

background in hall to minimum— guides to transport beams to an adjoining laboratory— focussing monochromators— position sensitive detectors— hot and cold neutron sources

Except for the hot and cold neutron sources, all other aspects

have been implemented and the throughput of the six spectrometers

in operation are at least 10 times and sometimes about 25 times

that of their counterparts at Cirus; going by the ratio of fluxes

alone at the centre of the pile, the throughput improvement

should have been only about 3 at the maximum power levels.

Efforts to build the hot and cold neutron sources as well as al

2D—detector SANS are underway.

Table IV lists the spectrometers currently in operation and

the principle and new types of investigations which we have been

able to carry out at the Dhruva reactor.

45

Table IV ^. Neutron Spectrometers at Dhruva and some typicalSolid State Studies (March.1993)

Spectrometers Solid State Studies1. Profile Analysis Powder

Diffractometer

2. Single CrystalDiffractometer

3. Triple Axis Spectrometer —

4. Polarised Neutron AnalysisSpectrometer

5. Hi—Q diffractometer

6. Filter DetectorSpectrometer

7. Small Angle Spectrometer

8. Medium ResolutionInelastic Spectrometer

9. Neutron Interferometer10.Spin—echo spectrometer

Thermodif f ractograms of mixedferrites to monitor magneticphase changesIn— situ loading of PdDStudy of hydridesUsed mostly for study of some50 different samples of Hi— TcmaterialsTrial texture studies of heattreated stainless steelcomponentsPhonon density of states ofTetracyanoethylene, YBCO,TICaBaCuO, ResorcinolParamagnetic phase studies of

Amorphous SejL-«Se«, H2O-DaOmixtures and glasses used forwaste retentionUnder installation

At Cirus :of micelles of CTAB at variousconcentrations, pH values andtemperaturesf errof luidsQuasielast ic scattering in NH-»salts (trial runs)At Cirus s Under relocationUnder development

It is because of increased throuhputs we are able now to measuredata like the phonon density of states, structure factors of

liquids and amorphous solids, polarised neutron analysis and high

resolution quasi-ealstic scattering.

3. Means of improving performance of spec trometers etc.

Performance of spectrometers can be improved by a variety of

means as has been documented in various publications. One can

resort to multi detectors, multi analysers, focussing

guides/collimators, focussing monochromators bent vertically and

horizontally as are already well known. But these demand high

technology and large finances. The other avenue open is by

46

increased use of the rather ubiquitious commercially available

rather cheap data acquisition, spectrometer control cum analysis

systems based on personal computers. The spectrometers operating

at Dhruva are all connected to dedicated indegeniously developed

on—line microcomputers which can control the instruments as well

as as carry out data acquisition and analysis. Conventional as

well multi—channel analyser data acquisitions are available.

Analysis such as peak-hunt, Rietveld analysis and display can 'be

routinely • carried out using these systems. In some special

situations like phonon measurements, one may like to assess

effect of resolution, which varies with spectrometer's various

angular parameters, even before measurements are carried out. We

have recently developed such a package in order to carry outtime—effective measurements on our triple axis spectrometer.

Comparative data displays are a must to ensure accuratenormalised data acqusition in many day—to—day situations

especially if one were studying samples of slightly different

compositions. Stable electronic systems are essential ingredientsof PSD based systems. Development of such systems based on

anologue or digital approaches using rise—time descrimination,

charge division or delay time methods are adopted at Dhruva.Recently our electronics engineers have designed and .developed

special ADCs which can act as ratio circuits also in amultiplexed manner to overcome difficult matching situations.

4. Optimum exploitation of indigenous resources

Neutron beam research program at Trombay has served verymany useful purposes and has influenced other activities in

physics research in general in our country. It was realised very

early in our development that a meaningful R&D program in solidstate studies via neutron scattering can be sustained if

47

complimentary developments and studies were also initiated andexpertise developed.

X—ray diffraction, Mossbauer spectroscopy and Laser RamanSpectroscopy were initiated within the research group; they haveblossomed into their own complete research activities over theyears. The protein crystallography activity which got startedwith meagre x—ray diffraction facilities has turned out to be oneof the most advanced schools in India. Thanks to the extensive

computer and graphics support available, the group has solvedrecently a protein-structure that of human carboxyanylase-based

on several activities involving isolation of protein from humanblood samples, protein crystallisation,. synchrotron x—raydiffraction data acquisition and analysis based on computermodelling. The one and only low temperature high magnetic field

Mossbauer spectrometer in the country designed and developed at

BARC is being routinely used for understanding magnetism of Hi—Tcmaterials. The Raman spectroscopy group has recently commissioned

a triple-monochromator based optical multi—channel system forinvestigation of binary and ternary semiconductors, hi-pressure

Raman spectroscopy etc.

Detector fabrication, instrument design and development, low

temperature accessories were all undertaken as a part of

inttntional indegeneous instrumentation activity. The aim was tomaximise indegeneous resource involvement and to minimise, if noteliminate, external inputs. This approach has paid very rich

dividends in terms of developing expertise in design,manufacture, machining, testing and commissioning not only all

the neutron spectrometers and related equipments but also other

equally important equipments like detectors, monochromators,

guides, cryostats, magnets, power1 supplies, RF units, control

48

systems, data handling systems etc. The fairly large variety ofinfrastructure available in BARC (in various disciplines) andfree flow of cross-disciplinary collaboration amongst various

disciplines has played a crucial role in developing proto-types

and even finished goods internally if necessary. Over the years,

some of the finished raw materials and some of the equipmentshave been got developed by commercial industries and other publicsector and private sector companies in the country also. Theexpertise developed helps in undertaking rather complicated

activities which result in technological fall outs in addition to

generate the capacity to operate and maintain our own systems. Weare also able to offer indegenous products like boron carbide,boron plastics, berylium blocks, control and data acquisitionsystems and neutron spectrometers on commercial basis to other

countries; we have done so bilaterally to Korea and UK andthrough the IAEA to Bangladesh recently.

As illustrations during the presentation of this paper I

intend to show via slides our activities related to manufacture

of spectrometers, development of guides and cold source.Therefore, indegenisation has not been merely a slogan as

far as we are concerned, it has become a reality which hasenhanced our confidence to take on more difficult tasks.

Currently, installation of a spin—echo spectrometer and a 2D

small angle neutron spectrometer have been occupying our interest

but there is great enthusiasm amongst my colleagues to go forreflectometry, neutron topography and other applications related

instrumentation. The only thing that is limiting our speed islack of enough trained manpower. Hopefully, we will overcomethis in the near future.

49

5- Requirements and Feasible^ Approach/es for Staff Training

Unlike in some of the advanced western laboratories where

the large facilities are manned by large technical manpowerrunning to hundreds sometimes and users also numbering a fewhundred as they come from various universities and other

institutions, our program is essentially based on in-house staffconsisting of about 20 scientists and 10 technical assistants.This total staff undertakes design and development of equipmentas well as operation and maintenance of the same as well as carryout research activities. The situation is somewhat like what usedto exist in English or European Universities with youngerstudents and senior professors.

Scientific staff are all M.Sc degree holders in physics andthe technical staff, science degree holders qualified in varioustrades like electronics or vacuum technology or laboratory

practices. The scientific staff have also undergone a one—yearadvanced course in physics in several topics not commonly coveredin Universities and also exposed to several experimental projectsin aur Training School. Such young and fresh staff are attached

to experimental groups carrying out solid state physics studies

at the reactor and are expected to carry out research from thenan. It is observed that, over a period of 3-4 years, most of them

pick up sufficient experience to be able to carry out experimentsreliably. They shape up as independent research workers over aperiod of 10—12 years. Opportunities are also created to enable

them to proceed on post-doctoral or equivalent assignments tooverseas laboratories in their assigned fields of specialisationto enlarge their expertise.

I have already referred to the small staff involved in the

neutron beam research. Increasingly it is 'felt that there is a

50

need to augment this by suitably trained manpower particularly atthe research level. An intentional effort was made about 4 years

ago when Dhruva reactor was declared as a National Facility forNeutron Beam Research. It meant that the facility would beavailable for researchers from the Universities and otherinstitutions and the BARC scientists would also take on theresponsibility of Contact and Guiding Scientists for approvedprojects. The University Brants Commission and the Department ofAtomic Energy created a consortium called Inter UniversityConsortium for Department of Atomic Energy Facilities (IUC-DAEF)and one of the Centres of the Consortium to use Dhruva was

founded. From then on, efforts have been made to train universitystudents and staff in techniques of neutron beam research viaworkshops and hands—on practicals as well as encourage them towrite proposals for experiments to be carried out at BARC. Four

workshops have been conducted so far on : (i) general techniques(ii) powder diffraction (iii) liquids and amorphous systems and(iv) small angle scattering. The next workshop will be related to•applications in texture, residual stress measurements and smallangle scattering in metallurgy etc. The Consortium supportstravel and local expenses for researchers, provides token supportfor chemicals and small equipments and expects that activitieslike sample preparation and characterisation by non—neutronicmeans are carried out at the Universities. It also encourages

Universities to design and develop ancillary facilities that canbe installed at Dhruva. It is expected that a few selected staffmembers of IUC—DAEF will also be stationed at Dhruva to

coordinate, help and train the University researchers as well ascarry out collaborative research programmes. They would be a veryuseful component as they can take over as Contact Scientists alsoreleaving the local staff for other activities.

51

6>. Regional and International Cooperation

We have already referred to the RCA activities in whichIndia was involved as a part of the Regional Cooperation amongstcountries in Asia. This activity in the field of NeutronScattering has somewhat diminished compared to that in thesixties. However, India continues to support this by holdingperiodical workshops on topics related to neutron scattering. Thenext one is scheduled to take place during November-December, 1993. Indian technical expertise can be made availablefor any activities in the region whether in design andfabrication of spectrometers or in providing research experts.

7. Possibilité es of indeqeneous and cost effective fabrication of

spectrometric components, instrumentation and guides

The spectrometers installed at Dhruva are all based onindegeneous design and development. An optimum mix of shield

materials and its shilding efficacy of the main monochromatordrum and its yoke were determined by shielding calculations.Except for the 1.2 metre diameter bearing that carries the mobile

monochromator drum load, all other components were fabricated outof materials commercially available in the country to tolerencesspecified. Rigid quality control to meet acceptability criteriaand international standards were maintained at every stage of

manufacture and assembly. In—house workshops as well asindustrial machine-shops were involved in the manufacture.

Electronics counting systems were obtained from theElectronic Corporation of India Ltd., and on—line computer

control systems from a state—owned commercial concern KELTRON.Detectors were fabricated in the division.

52

We have developed the guides indegeneously and it may beconsidered as a major developmental effort. Float glass plates ofrequired size (100x25x1 cm3) were imported and these were coatedwith nickel using a 3m dia high vacuum coating plant situated atoptical observatory at Kavaloor, India. 0.5mm dia nickel wire wasdrawn to our specifications by Midhani, tungsten filamentsdesigned at BARC were manufactured by Central Electronics Limitedand other mechanical fixtures made at our BARC Workshop. The

tungsten filaments had to be replaced after each coating run asthey would break because of brittleness as a result of alloyformation. The nickel coated plates were assembled into 1m longrectangular guides using precision jigs fabricated at BARC. The.

fixtures to align the guides, vacuum enclosure, mechanicalsupport etc., were all obtained from indigenous sources» vacuumtested and installed in position using optical instruments.

8. Commercial and Industrial Applications

We have not carried out any studies of commercial orindustrial applications so far. However, we are in the process ofdeveloping techniques of texture and residual stress measurementson our diffractometers for testing mechanical components used in

our reactor systems. We had carried out texture determinationsome time earlier to assess quality of uranium metal fuel rods atthe stage of developing the process.

Next page left blank 53

SOLID STATE PHYSICS AROUND PAKISTANRESEARCH REACTOR AT PINSTECH

N.M. BUTT, J. BASHIR, M. NASIR KHANPakistan Institute of Nuclear Science and Technology,Islamabad, Pakistan

Abstract

The role of PINSTECH has been to conduct research and development asnecessitated by the programmes of the Pakistan Atomic Energy Commission. Thisassignment was entrusted to PINSTECH from its very inception in early 1960's. Thesetup of PINSTECH and its facilities were planned at this time and the first majorresearch facility in the form of 5MW swimming pool research reactor was establishedin 1965. This reactor went critical in December 1965 and attained full power in June1966. The emphasis of course was on the utilization of the research reactor for thestudies of nuclear physics, solid state physics, activation analysis and for the productionof radioisotopes. The main facility for the study of solid state physics is the triple axisneutron spectrometer. In this presentation, after giving a brief description of theexperimental facilities in and around the reactor, the utilization of the reactor for avariety of research projects in the field of solid state physics is described. A briefdescription of the recent upgradation of the reactor power to 10 MW is also given. Atthe end, planned future activities suited to the available thermal neutron flux (about8xl013 n /cm2 sec) are reported.

INTRODUCTIONNeutron scattering is by now a well established technique and with the advent of

nuclear reactors as a source of neutrons a wide variety of phenomenon have beenstudied. The usefulness of thermal neutrons arises from its special properties like itslack of charge, and its magnetic moment. As the wavelength of thermal neutrons iscomparable with the interatomic distances in a solid, neutrons diffracted from thecrystals provide information on their structure. Also the energy of the neutrons iscomparable to that of phonons, the inelastic scattering of neutrons provides an accuratemethod of determining the energy of phonons and hence the frequency of latticevibrations. Its lack of charge implies that neutrons can penetrate several centimetersdeep into the material and hence bulk of a material can be studied.

The Pakistan Research Reactor (PARR-I) was established in 1965 with the aimto promote the technical know-how that is necessary for the introduction of nuclearpower in the country. With the establishment of 5MW research reactor, a neutrondiffraction facility was indispensable. Hence a triple axis neutron spectrometer wasinstalled at PARR-I. With this medium flux reactor and a simple triple axisspectrometer, a variety of projects such as phase transitions, texture, lattice dynamics,crystal structure, and measurement of temperature factors of materials have beencompleted. The utilization of this reactor in a developing country with limitedinstrumental facilities leading to research publications of international recognition is asource of encouragement for other developing countries of the region having similarsituations.

This instrument operated successfully until recently when its was decided toupgrade the reactor power from 5MW to a nominal power of 10MW. With the

55

upgradation of the reactor, it is planned to install new neutron diffraction facilities sothat the reactor can be utilized more effectively.

NEUTRON DIFFRACTION STUDIESFor the neutron diffraction studies a triple axis neutron spectrometer (TAS) was

acquired from Poland. The spectrometer has been installed at the beam port 3 andmoves on rails such that it can be moved up to 25 feet away from the reactor wall.

The TAS has 5cm x 5cm beam aperture. Using the soller collimators thecollimation at the monochromator and analyser system can be varied between 10' to60'. The overall neutron background using a 60 cm long, 5 cm diameter BF3 detectoris about 2 counts/minute. The monochromated neutron flux at the sample position isabout 105 n/cm2/sec at a wavelength of 1A- The main features of the spectrometer aregiven below.

- ci

01

Moiiocliroinatoc

Fission Chamber

Sumpfe Table

AnalyzerBI73 Detector

Sollcr collimators

Fig.l Layout of the Triple-axis Neutron Spectrometer TKSN-400 at PARR-I

56

MAIN FEATURES OF TRIPLE AXIS NEUTRON SPECTROMETERTKSN-400

The layout of the triple axis spectrometer is shown in Fig. 1. The mechanicalparts consist of the first-axis where a single-crystal monochromators (M) is placed. Thecrystal is surrounded by a very heavy monochromatic shield of about 3 feet radialthickness of borated water. The crystal rotation table and the monochromator arm arecoupled through gears in the ratio 1:2. The table has an angular range of rotation of360° while the monochromator arm (or shield) can rotate over a range of -15° to +90°.The accuracy of the angle setting is 1' and the backlash in gear coupling is about 2'.The second-axis system, the sample table (S) and the sample arm are placed on themonochromator arm. The sample table has no gear-coupling with the sample arm andboth can be rotated independently. The angular ranges of the two are 360° and 90°respectively. The third-axis, analyser table (A) and the analyser arm are placed on thesample arm. The angular range of the table is 360° while that of the analyser arm is160°. All the angles of all the axes can be set automatically by a programmed papertape. This spectrometer has been used for the following studies: -

(a) ELASTIC NEUTRON DIFFRACTION(1) DEBYE-WALLER FACTORS OF MATERIALSThe intensities of x - rays or neutrons scattered elastically into Bragg diffraction

peaks by crystals is proportional to a Debye - Waller factor exp(-2M) which accountsfor the thermal disorder of the atomic motion.Therefore, the observed intensity, IQ, ofa Bragg peak is written as

I0 = Icexp(-2M)For a monoatomic cubic crystal M is related to the thermal parameter B by

2M = 2B(sin6/A,)2 (1)

where A, is the x- ray or neutron wavelength, 0 is the Bragg angle. The B parameter isrelated to the mean square amplitude of atomic vibration (u2) and the Debyetemperature 0 of the crystal through the relations

B =8*;2<U2> (2)

= (oh /̂mk ©2 )[ <£ (x) + x/4] (3)

where the symbols have their usual meanings (James 1967).

MEASUREMENT PROCEDUREThere are two different methods of determining the B values of the crystals, viz.:1) From the variation of integrated intensities with Bragg angle (Wilson plot)2) Rietveld refinement procedure

WILSON PLOTIn this method, intensities of a large number of Bragg reflections are measured

at a fixed temperature. The integrated intensity for each reflection is then determined

57

by summing the counts, point by point, across the Bragg peak and subtracting thebackground on either side. These integrated intensities (lyj) are converted to thestructure factors ( F )using the standard relation

= C hkl

with C is the scale factor, J^ the multiplicity of the reflection, Lj^ itsLorentz factor of the (hkl) reflection. The temperature parameter B is then determinedfrom the least square fit to the line

ln(F0/Fc) = Constant - B (sin 0A )2 (4)

The slope of the line [Fig. 2] gives B at the temperature at which the diffractionpattern [Fig. 3] is measured and the mean square amplitude of vibration (u2) and theDebye temperature © are calculated from equations (2) and (3).

0 0.2 0.4 0.6 O.S J.O 1.1 l.-l u, l.j; J u .\

Fig.2 Wilson Plot for Al (X-ray diffraction data).

In case of cubic compounds, there are two atoms and hence two different B-values, B+ for cation and B" for anion. In this case, B determined from (4) gives theaverage B value which is the mass weighted average of B+ and B".

It is also possible to determine the individual temperature factors of cation andanion from the diffraction experiment. For example, in case of ZnTe which has zincblende structure (space group F43m), the structure factors can be expressed in threeforms of the sum of h, k and 1 related to 4n, 4n±l or 4n±2, n being an integer. WithZn atoms at (000) and equivalent positions, and Te atoms at (1/4,1/4,1/4) and equivalentpositions, the structure factors are given by

58

= 16 [ + bTe exp(-MTe)p for h+k+1 = 4n (5 )

4u±i = 16 [ bz,, exp(-MZn)]2 + 16 [bTe exp(-MTe)]2 for h+k+1 = 4n ± 1 (6)

= 16[bZnexp(-MZn)-bTeexp(-MTe)]2 for h+k+1 = 4n±2(7)

where bZn and b^ are the neutron scattering lengths of Zn and Te and n is positiveinteger. From (5) and (6) we can write (Butt et al 1978)

(200 >

mol

2»——— — RKAGG ANGLE ( DEGREES! — -

Double-axis neutron diffraction pattern of KC1 powder.

, 500C

Fig.3

JO Ji )J

ÜRAGO /.HOLE | 0 £ G K t l S | — — -

Triple-axis neutron diffraction pattern of KC1 powder.

A typical Neutron Diffraction pattern in Double axis and Triple axismode.

59

ln(a+ß) = Constant - B^ (sin 9A, )2 (8)In(a-ß) = Constant - BTe (sin 6/X )2 (9)

where

ß=[2F*4n±1-F>4nF2 dl)

Thus the plot of ln(a+ß) and ln(a-ß) vs (sin 0A,)2 will be straight lines; the slopes ofthese two straight lines yield BZn and Bj,. respectively.

CORRECTION FACTORSThe correction factors include the contributions of thermal diffuse

scattering(TDS) and absorption.

THERMAL DIFFUSE SCATTERING (TDS)In the diffraction experiment, the scattered intensity consists of both elastic and

inelastic components. In fact the elastically scattered component is measured on abackground of inelastically scattered radiation consisting of a Compton - scatteredcomponent which has a slow variation with the scattering angle, and a phonon -scattered component which is known to be peaked around the Bragg position and theprincipal part of which is proportional to M. The peaked part of the inelastic scatteringcannot be completely resolved from the elastic Bragg peak owing to insufficient energyresolution of the experiments. The inclusion of this part of the phonon scatteredradiation in the measurement of the Bragg peak intensity thus results in a systematicerror in the estimates of the temperature factors.

This systematic error can partly be corrected by the use of triple axisspectrometer whereby the resolution is improved as compared to the two axisspectrometer. In the triple axis spectrometer, a second single crystal is placed after thesample. This single crystal, the so called analyser crystal diffracts only those neutronsto the detector which have the same energy as those of incident neutrons. In this way,inelastic contribution to the Bragg reflection is considerably reduced and therefore moreaccurate B values can be determined.

The TDS contribution to the diffracted intensity can be estimated theoretically[Nilsson (1957), Warren(1953), CMpman & Paskin (1959), Willis(1969)]. If Iobs is theobserved integrated intensity, the TDS corrected intensity is given by

I . = !„ (1+ a) (12)obs Bragg

wherea = (871/3) (Q2qKbT/3mv z)[ l/V,2+2/Vt*] (13)

where the symbols have their usual meanings (Willis 1969). Eq. (13) shows that a isproportional to Q (for a fixed value of q^). If a is sufficiently small, then (1+a) canbe replaced by ea then the effect of including TDS in the estimate of the Braggintensity is to decrease artificially the overall B-value. This decrease AB^s is given by

ABTDS =(647t3/9)(qmaxKbT/mvz)[l/V12+2/Vt2] (14)

60

Willis (1969) also pointed out that the first order TDS correction for the faster thansound neutrons can be estimated from the same formulae as for x-rays, and that, to firstapproximation, there is no TDS correction for the scattering of slower than soundneutrons.

ABSORPTION CORRECTIONFor materials with high linear absorption coefficient, the change in B due to

absorption is (Hewat 1979)

ABabsor = X2 d(flr) + d(|ir)2 (15)

where ]J. is the linear absorption coefficient and r is the radius of the sample. The valuesof the two constants dj and d2 for the cylindrical sample are -0.0368 and -0.03750respectively.

RIETVELD REFINEMENT METHODIn the Rietveld method of refining powder diffraction data, the profile of the

entire diffraction pattern is calculated and fitted to the experimental profile. There is noneed to extract integrated intensities first and so patterns can be analysed containingmany overlapping Bragg peaks. The method was applied originally by Rietveld (1967,1969) to the refinement of neutron intensities recorded at a fixed wavelength.Subsequently, it has been successfully used for analysing powder data from all fourcategories of experimental technique, with neutrons or X-rays as the primary radiationand with measurement of the scattered radiation at a fixed wavelength (and variableangle) or at a fixed angle (and variable wavelength).

A structural refinement based on powder data is always likely to be inferior toone using good single-crystal data. In cases where it may not be possible to obtain asingle crystal sample, with a powder sample twinning, absorption and extinction maybe neglected. The Rietveld method is suitable for refining structures with up to 200parameters: for a review of successful refinement, see Cheetham and Taylor (1977) andHewat(1985).

The function Mp which is minimized by least squares is

Mp = £• W. [{yi(obs)-bi} - yi(calc)]2 (16)

Where yi(obs) is the intensity measured at a point i in the diffraction pattern,w^ is its weight and bj is the contribution from the background. If the variance of thebackground is arbitrarily set to zero, and if the only source of error is that fromcounting statistics, then

Wi = [yi(obs)j-lyi(calc) is the calculated intensity and is given by

61

in which 1^ is the integrated intensity of the kth reflection whose peak shape isdescribed by the function Gjk which is normalized so as to give a value of unity whensummed over the whole peak. The summation in (17) includes just those reflections,making a significant contribution to yi(calc); the summation in (16) is over the number,N, of points used in the refinement.

In the fixed wavelength neutron technique, the intensity is measured at differentscattering angles 29j. The integrated intensity 1^ is then:

Jk = c JkLkFk (18)

with C is the scale factor, J^ the multiplicity of the reflection, L^ its Lorentzfactor and F^ the structure factor which is calculated from a starting model of thestructure; for magnetic structures there is both nuclear and magnetic scattering, and so(for unpolarised neutrons) F^ is given by

[Fkl2= [F.JJ2+ [Fmk]2 (19)

where Fnk and Fmk are the nuclear and magnetic structure factors respectively. Thepeak-shape function is approximated by a Gaussian

Gik = [HkH exp [ -4In2( 2^ - 20k )2/H2k] (20)

and Hfc is the full width at half maximum and varies with the Bragg angle 26 inaccordance with

Hk2 = Utan2 6 + Vtan 0 + W (21)

where U, V, W are constants and depend on the collimators and mosaic spread of themonochromator(Caglioti et al, 1958). The contribution of the reflection k to yi(calc)effectively drops to zero when ( 20j - 20k ) exceeds 1.5 Hjj.

From the least-squares refinement, information about the thermal parameterscan be obtained.

We have measured, the Debye-Waller factors of various elements and alkalihalide materials. Furthermore, concentration dependence of the Debye temperature inmixed alkali halide systems was studied. For these systems, it was observed thatconcentration dependence of Debye temperature is not linear. The results also indicatedthat the values of mean square displacement of atoms obtained using the diffraction datawere composed of two factors. One due to the dynamical factor based on latticevibrations and the other due to the static variation in atomic positions. The staticvariation would occur due to different sizes of the constituent ions. Knowledge of thesefactors is important and one has to be careful in evaluating various constants likespecific heat, elastic anisotropy, shear and Young's moduli for these materials. Table 1summarizes the results of these investigations.

Reliable values of temperature factors are required in the calculations of TDS,EXAFS, LEED, impurity scattering and band structure. These are further useful forreactor moderator studies and for the test of lattice dynamical models used tounderstand the binding forces in the solids.

62

Table 1: List of materials for which thermal parameter B has been determined atPINSTECH by the powder neutron diffraction method.

Material B+(A2) ß-(AZ)

MoK0.5Rb0.5C1

SiRbClNbKFRbFKasRbo-S17

T1C1RblKI" 0 7 0 3JVA 5-KDfi cJL

Ko.aRbo.?1

ZnTe 1.26(5) 0.74(5)UC>2 0.23(77) 0.43( 7)KBr 2.55( 7) 2.20( 4)

B(A2)

0.25(1)2.39(12)0.45( 2)2.43(20)0.55(5)1.23(11)1.40(25)1.84(20)3.07(22)3.19(11)3.06(16)3.10(24)3.52(17)3.14(11)0.91( 5)0.25( 9)2.33(30)

0(K)

385(7)172(10)531(11)153(7)262(12)312(14)216(19)214(12)97(4)101( 2)117(3)112(4)102( 3)105( 2)198( 5)396(70)158(4)

Reference

Bashir et (1992)Bashir et al (1992)Beisheng et al (1990)Ghazi et al(1989)Bashir et al (1987)Beg et al(1981)Beg et al (1981)Beg et al (1981)Mahmood et al(1980)Beg et al (1979)Beg et al (1979)Beg et al (1979)Beg et al (1979)Beg et al (1979)Bashir et al (1988)Ahmed et al (1979)Butt et al (1976)

Although temperature factor data for materials is available in the literature, butin some cases large variation in B-values are observed. A typical example is that ofGaP where average temperature factor ranges from 0.38(3) to 0.74(3) A2. Therefore inorder to provide reliable values of temperature factors, the same were compiled andexperimental data was evaluated. From the evaluated data, recommended values oftemperature factors of 22 cubic elements [Butt et al, 1988] and 52 cubic compounds[Butt et al, 1993] were published. Furthermore using these data, interesting correlationsbetween the thermal parameter B with various thermal properties [such as melting point(Fig. 4 & 5), coefficient of thermal expansion (Fig. 6)], mechanical properties [suchas Young's modulus (Fig. 7) and compressibility (Fig. 8)] and cohesive energy (Fig. 9)of cubic elements have been established (Butt et al 1993). Similar relations have beenobserved in case of cubic compounds (Fig. 10 & 11).

(2) NEUTRON DIFFRACTION STUDIES OF THE UNIT CELL OFCELLULOSE - 7, CELLULOSE - HAND DEUTRATED CELLULOSEThe unit cell of cellulose-I has been a subject of study for about 50 years using

x-ray and electron diffraction techniques. The most widely favoured structure ofcellulose-I has been the structure of Meyer and Misch (1937) who gave «=8.35A,Z>=10.3A c=7.9A and ß = 84° based on their x-ray work. However, it was pointedout by many workers (Jones 1958, 1960) that the structure was not completelysatisfactory. They argued that agreement between intensities calculated on the basis ofthis structure and observed intensities is not good. Furthermore, it is not possible to

63

100

10

*<^»'

aï

0.1

K

i Na

Observed Calculated

100 1000 10000Tm(K)

Fig.4 Correlation of Thermal parameter with Melting Point Tm. (CubicElements)

0.001

Fig.5

0.01a2 /T (A 2K

Correlation of Thermal parameter with a2/Tm. (Cubic Element).

64

100

10

0.1

Observed —— Calculated

Fig.6

100 r

Na»

Li*

. « A l

Ta'"Mo.

10 100

Correlation of Thermal parameter with coefficient of Thermalexpansion (Cubic Elements)

10

0.10.01

Fig.7

Observed —— Calculated

Nb - I r

W

YMO" 6 ( K g / c m 2 )Correlation of Thermal parameter with Young's Modulus. (CubicElement).

10

65

100

• Observed Calculated

10

NiW

0.11 10 XnoW/Kg)Fig.8 Correlation ofThermal parameter with compressibility. (Cubic

Elements).

1000

100

Observed Calculated

10 •K

- N bAu

0.10.1 1 10

U«(eV)Fig.9 Correlation of Thermal parameter with Cohesive Energy Uc . (Cubic

Elements).

66

10

CJ

ÛÛ

0.1100

Fig.10

OBSERVED CALCULATED

CUCI'

ZnSe

RbBfKC\

GaA»SaP

ri XNp ; \>

ZnS .GaSbSrF2

BaF

UCi

«MflO

TiC

ioooT m (K)m

10000

Correlation of Thermal of Thermal parameter with Melting Point(Cubic Compounds).

10

CSJ

m

0.10.1

Fig.ll

OBSERVED OBSERVED

RbC

NaCI

N&C

1 11 10 2Y*10 (dynes/cm )100

Correlation of Thermal parameters with Youngs Modulus (CubicCompounds).

67

index all the reflections on the basis of Meyer and Misch unit cell. Honjo & Watanabe(1958) carried out the examination of cellulose - I by low temperature electrondiffraction technique and reported that a number of reflections can be indexed, not byMeyer and Misch unit cell, but by a cell having a and c dimensions twice as long asthose suggested by Meyer & Misch.[Fig. 12 ]

Neutron diffraction studies conducted at PINSTECH gave the unit cell forcellulose -1 as a = 16.78A, b = 10.3A, c = 15.88A and ß = 82° [Beg et al 1974]andhence we are able to confirm the results of Honjo & Watanabe. We were also able toobserve the diffraction peaks which were not observed in the X - diffraction pattern.The studies were latter extended to cellulose-II [Fig. 13]. These experiments wereperformed with plate as well as cylindrical samples to avoid the preferred orientations.The cell dimensions were again found to be larger than those given by x - raydiffraction technique and were a = 15.7A, b = 10.3A, c = 18.4A and ß = 63°[Ahmed et al 1976].

(3) ORDER-DISORDER PHASE TRANSITION STUDIES IN IRON BASEDALLOYS

In iron rich Fe-Al alloys, two ordered structures with stiochometric compositionFeAl and Fe3Al can be formed from the disordered bcc alpha - phase. The FeAl phaseis formed by ordering among the nearest neighbour atoms of the alpha phase to formthe CsCl type structure with iron atoms at the cube corners and at the body centers. Afurther ordering among the second neighbour atoms produces the Fe3Al structure witha unit cell having twice the lattice parameter of the phase.

The order-disorder transition in FeAl alloys was first reported in 1932 byBradley & Jay. Since then, much effort has been put into understanding the order-disorder transformation in different systems particularly in the FeAl system. Howeverthere is some disagreement with regard to the type of transition from disordered toFeAl, FeßAl ordered phases. According to some workers, this transition is a first ordertransition, while according to others it is of a continuous nature.

The neutron diffraction technique provides a direct way of investigating theorder-disorder transition by studying the intensities of the superlattice peaks [(111) and(200) in the present case] in the diffraction pattern. The intensity of peaks dependsupon the degree of the order and on the difference between the scattering lengths of theatoms involved.

We studied the order - disorder phase transition in Fe Al alloy for compositions20.24, 24.15, 28.06, and 31.45 at. % Al.fAhmed et al 1982] We confirmed that thephase transition from the disordered alpha phase to ordered FeAl and from FeAl toFe3Al is of continuous nature. The value of critical index ß, the order parameter, wasfound to be 0.302(9) thereby indicating that the transition is the first order transition.The phase transition in sample with 24.15 at. % Al is shown in Fig. 14.

(4) TEXTURE STUDIES IN SHEETS OF COPPER ANDALUMINIUM

The great majority of technologically applied materials have a polycrystallinestructure. In such materials the texture describes statistically the orientation distribution

68

bun i— n —— • ———— ; ——————— r— — ••IOC L |M* • '

ft '200 L 1 ', i ,T-19,000 •- ' ' ' i

800 - ' i 1 II 1 '6 00 - l' 1 Hltt |i o ° - M I j u ,200 - J ' d .

vi 16 ,000 - 1 i {'*'§ 800 -ou 600 -

g iOO -

Z 1 7 . C C O -800 -SCO -400 -

200 -15,000 -

SOO -500 -

I B ; O O i i i i 1 i i i i ! i i 1. 1 ..itO SC 50 70

, ; :, : • , 1 , i , , . . . i

101 ( M M )i

( 10Î ( M M )

Î • 1 ®

• 1 • 1 T—TT — n — r— r IT j i : — i—i —-

----

| . ! ' , !•_ 002 ( M M )

/ 'j1 t ' ^ " ^ 1 * 1 4

1 ' 1 • 1 1 1

1 5 1 1 i 1 ittl i j l l . , ,.,''22 1 1 U ,'4,li,yÏ02 * U U l ' 1

L

l^l' 21Î 2 Î

. ,' • , 1 , , ! . ! , . . . 1 I ! ! ! 1 i

80 90 IM } 1IC,SIN c/j, x l O . A —

:1 1! ' :'f'|i1 'ii ' !** '1 'Î I t l

1 11 !1

, , ! , . , ,

1

1, ,|)i 1 t

' n in'|til|j]

. i . . i .120 130 UÛ

-_

-

^'-'

w i ih i iL'iii'lJJll*.1 *' 'i1 i' ri

i, , i , , . . iISO 160

Fig. 12 Double-axis Neutron Diffraction pattern of Cellulose-L

CÇ i —————————————————————————————————————————————————————————————————————————————————————————————————————————————————

SB

54

53E:

3

1 SO5 is- is

~

:

"

-

SO• i 1010)

«V.V\\

H|H

s-c. ac1202) (OQ3I(200!

^ 'i 1 ' ' i jl M M » ' M » t H ^

( 1 0 1 )t• X - R A Y

( R e f . 5 )

S-C

, »

, » ' M * !i '

Single Cel l. Doubt» Cell

t

' *»" 1 J H H'Si1 ^

( 1 3 T ) ( 002)i i

X - R A Y„,,, X - R A Y( R t ( - 5 )

5C S; EC 65 70 75 60 S5 30 35 ICO 105 HO US ::0 125Q 3 • -'

S. 1 x 1C A '

liO

Fig. 13 Neutron Diffraction Pattern of Cellulose-II.

69

6OO

5OO

4OOO

Z 3000u

2OOO

1000

o

—— >-(200)

ö o

J l5OO 54O GOO 6<10 7OO 7'10

TEMPERATURE (°C)

Fig.14 Temperature dependence of the (111) and (200) superlattice peakintensity in Fe-Al (24.50% Al).

of crystallographic planes of the grains (crystallites) with respect to set of macroscopicsamples axis i.e. the texture gives information about how large a volume fraction of thematerial having a specific crystallographic orientation but no information about wherein the material these different grains are positioned. When the texture is strong,, mostof the grains have almost identical orientations, whereas, when the texture is weak, thegrain orientations are almost random.

Texture may have large effects on materials properties, for example on strength,anisotropy, ductility, fracture behaviour, fatigue resistance, and magnetic properties. Incharacterization of metallic uranium fuels the texture plays an important role. It isimportant to know the dimensional changes due to prefered of these changes on thetype of orientation in the uranium sample. From an industrial viewpoint it is therefore,important to have a precise knowledge of the texture. Furthermore, texture is offundamental interest as it can be used as a probe to study metallurgical processes: mostmetallurgical processes, like plastic deformation, recrystalization and grain growth areorientation dependent, which means that texture commonly changes during theseprocesses. By measuring the change in texture, information about the crystallographicprocesses taking place in the bulk material is obtained. This type of data is not directlyaccessible through other measuring techniques for example by x-ray diffraction.

70

At PINSTECH sheets of 99.999% aluminium and copper were coldrolled to 92% reduction in thickness for texture studies by neutron diffraction. Theexperimental setup is shown in Fig 15. A piece of copper single crystal was also rolled.The results showed that all samples of f.c.c. metals did not attain standard texture. Oneof the aluminum samples ended up in [200] (002) texture (Fig. 16), whereas the sheetmade from copper single crystal followed the orientations of the original crystal. Onlythe specimen prepared from a billet which was carefully heat treated and mechanicallyworked gave standard orientations [Beg et al 1985].

(5) STUDY OF SUPERIONIC CONDUCTORSSuperionic conductors are materials of latest interest for their possible use in

future electric (vehicles) batteries. Superionic materials are the materials which undercertain conditions have higher ionic conductivity than those of molten salts andelectrolyte solutions. In CaF2 type materials, the transition from insulating toconducting state starts gradually with temperature. A large proportion of halogen atomsleave their regular sites and assume interstitial positions. The mobile atoms areresponsible for high conductivity of the material. The phenomena can be studied withneutron diffraction temperature whereby physics of the continuous order-disorder

R E A C T O R

MONOCMROMATOR — - •—,

SAMPLE

COLLIMATOR

DETECTOR

( a )

( b )

Fig. 15 Schematic diagram of Texture. Studies of Rolled sheets by NeutronDiffraction.

71

transformation can be studied. We have studied the variation of (111), (220) and (200)diffraction peak intensities in SrCl2 up to 800°C. A continuous transition is observedbetween 650 °C to 700 °C.

.n TO 60 $0 U» 30 20 10 — •<•—W 70 30 40 50 60 10 tu 50TOU——»

(200) po le Maure o f c o l d - r o l l o d a l u m i n i u m shoo t (s.vrole 1 1 ) .

... • . . . . . - *. . : ; / / / N'

90 go 70 «0 50 10 JO JO 10 — -< -- K) 20 30 to SO 6J ;u «D iO

(200) pole Mguro of c o l d - r o l l a d coppor s h o o f s l .

Fig. 16 Pole Figures of cold rolled copper sheets.

72

j l" innsverse optic

' 751-

Q 02CUWÛÔW 03 05 0,', 07--[wo] M- -

rcduci'd iwiv? vc

Fig.17 Phonon dispersion in mixed crystal Ko.sRbQ.sI.

(b) INELASTIC NEUTRON SCATTERING

LATTICE DYNAMICS OF K05Rb0J, K0jRb0.5Cl AND COPPER NICKELALLOYS

Studies of mixed materials and alloys are very important in the fields of sok'dstate physics and metallurgy. Alkali halides are ionic salts and they are easier to alloy.These materials are also easier to handle mathematically. First detailed latticedynamical study of mixed alkali halides was carried out at PINSTECH. Acousticphonon branches, both longitudinal and transverse were measured in a single crystal ofKO sRbo.sL (Aslam et al 1976). The research on these materials gave the followingimportant results:-

(i) Interatomic force constants and effective ionic charges are reduced onmixing.

(ii) The mixtures are elastically more anisotropic than the constituents,(iii) All mixed alkali halides have split or double phonons which dispersetogether. This is a direct result of mass disorder. The present work clarifiedmany misconceptions, about mixed materials, based on the infrared work.Fig.17 gives the dispersion relations and the shell model fit for Kg 5Rb0 5I. Thework was extended to a copper-nickel alloy single crystal. The above effectswere also observed for this alloy.

73

Scmple

Monitor counter

Seller slits

BF3 detectors

Soller slits

Cadmium masksBeom fromReactor

Germanium single crystal

Fig. 18 Lay out of a High Resolution Neutron Powder diffractometer.

FUTUREso far we have discussed the facilities available and the physics research that has

been done when the reactor power was limited to 5MW. In 1992, the reactor powerwas upgraded to about 10MW. This would result in approximately doubling the fluxthat was available before. This would help the experiments to be completed in less timeand with more accuracy. In future it is purposed to established the following newfacilities around the reactor.

HIGH RESOLUTION POWDER DIFFRACTOMETERThis instrument will be used for the measurement of residual stresses in the

materials and will be installed at the beam tube No.4 of PARR-I. The layout of thediffractometer is shown in Fig. 18. Disk shaped Ge single crystal will be used asmonochromator. Ge crystal is especially needed when second order contaminationmust be avoided. It will be adopted in the strain scanning mode by the addition ofautomatic sample scanner giving three orthogonal x, y, z translations and a rotationabout a vertical axis. The shape of the gauge volume will be defined by cadmiummasks. Positioning of the samples will be achieved using fixed theodolites.

NEUTRON RADIOGRAPHY FACILITYNeutron radiography technique is based on the principle that the workpiece

(sample) is exposed to a beam of thermal neutrons extracted either from the reactorbeam port or from a neutron source. The transmitted intensity of neutron beam throughthe sample exposes the converter screens/foil which then emit radiation sensitive to adetector (Fig. 19). The detector could be a 1) X-ray film, 2) SSNTD detector or 3)video camera with TV screen. The shadow image formed on the screen/foil is then

74

-cr>——— ^

—I to

-:•----- is

ä)\ ~ — 50J - - -

\\

~V~i :

S _J

\V

1 Guide Channel2 Reactor Core3 Trolley Wheels4 Shuttter Wheels5 Trolley6 Neutron Streaming Stop7 Beam Collimators8 Bismuth Filter9 Source Block10 Beam Tube No. 611 Beam Shutter12 Neutron Beam Trap13 Trap Support Structure14 Enclosure Walls

Fig. 19 Lay out of a Neutron Radiography apparatus at PINSTECH Reactor.

transferred to X-ray film by placing the film in close contact with the screen/foil. Theexposed films are then processed to reveal defects and flaws present in the sample.

This technique has an advantage over other techniques because of the randombehaviour of the neutron cross - sections for different materials. The attenuationcoefficient of different materials increases with increase of the atomic number butneutron cross - sections show a random behaviour. Due to this behaviour of neutroncross - sections the neutron radiography technique is being used to examine both lowand high atomic number materials. It can also be applied to active as well as non activenuclear fuels and components. Further advantage is the large penetration power ofneutrons in such materials which enables non-destructive testing of several centimeterthick samples.

An increase in the reactor power would almost double the thermal neutron fluxavailable at the sample. This would help reduce the exposure times.

This technique has applications in1) Nuclear industry2) Aerospace3) Explosives4) Archeology5) Materials analysis

SILICON TRANSMUTATION DOPINGThe use of silicon as a semiconductor material has increased rapidly and the

present world wide consumption is now excess of 6000 tons per annum (Meese 1983).Silicon is used in a wide range of electrical and electronic devices and the applicationsrange from high voltage thyristors to power diodes and transistors in conventionalelectronics and as the base material for integrated circuits used for computer systemsand microprocessors.

The most commonly used dopant is phosphorous for n-type silicon.Phosphorous doping can be achieved by the incorporation of the dopant in the moltenstage used during the float-zone method of crystal formation. However, this methodleads to an inhomogeneous distribution of the dopant and resistivity variations are ashigh as ±30%.

The application of a neutron transmutation doping process which shows muchsmaller variations has been of interest to silicon manufactures.

By the capture of a thermal neutron an atom of Si can be transmutated into theunstable isotope Si which decays with a half life of 2.62 hours by the emission of beta-particle to the stable isotope P.

30Si (n,y)31Si ____________ 31P

By means of this reaction, phosphorous atoms which act as electron donors canbe produced in the silicon crystal lattice. Thus the technique results in uniformallydoped n-type silicon with accurately predetermined resistivity values.

SMALL ANGLE NEUTRON SPECTROMETERNormally small angle neutron spectrometer (SANS) use cold neutron sources

and high flux reactors for the investigation of defect structures in materials. A newconcept using thermal neutrons instead of cold neutrons has been developed recently

76

Cu - PfïEMOHOCHROMAWK

SAMPLE

CHANNEL-CUTCRYSTALS

Fig.20 Lay out of a Small Angle Neutron Spectrometer (SANS) usingthermal neutrons.

(Rauch 1991). This small angle spectrometer [Fig. 20] is equipped with two monolithicchannel crystals increasing the angular resolution almost two orders of magnitude betterthan for other small angle instruments. In this way the resolution is decoupled from thedivergence of the incident beam. Therefore a divergent beam can be used on the samplethereby providing sufficient thermal neutron intensity available at the small andmedium flux reactors. This instrument is of versatile use, for example, in the study of;ageing and fatigue effects, precipitates, cracks, and other inhomogeneous ties inmaterials. The instrument has promising use in the study of materials science and willenable the opening of another type of studies of materials using neutrons.

ENGINEERING STRAIN MEASUREMENTS BY NEUTRON DIFFRACTIONTECHNIQUE

Heat treatment, welding and plastic deformation during fabrication can leavestrong residual stresses locked in within components. These stresses can affect, thecomponent life in service as they may add to applied loads causing fatigue and failure.

The presence of residual stress in engineering components can affect theirmechanical properties and structural integrity. Neutron diffraction is the onlymeasuring technique which can provide spatially resolved non-destructive strainmeasurements in the interior of the component. By recording the change in interplanerspacings elastic strains can be measured for individual lattice reflections.

Engineering strain measurement necessitates positioning of samples toaccuracies of 0.1 mm relative to precisely defined neutron beams. Different positionsare examined by translating the specimen through a defined volume. The strain indifferent directions is measured by re-orienting the specimen with respect to thedetectors. Thus the main requirements are a multi-axis manipulation system andprecise, repeatable collimation of both the incident and collimated beams.

77

The diffraction principles for the determination of strains are essentially the same forboth X-rays and neutrons when a beam of incident neutrons or X-rays is incident upona polycrystalline material whereby both scattering and absorption occur. The scatteringis not isotropic but is concentrated in "Coherent Bragg reflections" in directions definedby the Bragg equation

A = 2dsin0 (22)

where 20 is angle through which the beam has been scattered, d is the interplanerspacing and A, is the wave length of scattered radiation. A change in lattice spacing odwill, for a fixed wavelength produce a corresponding change 09 in the Bragg angle 6,is given by

59 = - tanG 8d/d (23)

where od/d represents the lattice strain s. It may be expressed as

e = od/d = -06 cote (24)

where the strain measured in that in a direction perpendicular to the reflecting planesand parallel to the scattering vector Q which bisects the angle between the incomingand scattered beams. Strain distribution through out a component may thus in principlebe determined by measuring 00 for neutrons scattered from different locations withinthe component.

STRUCTURE DETERMINATION OF HIGH Tc SUPERCONDUCTORSSuperconductivity is one of the most exotic phenomena observed in some solids.

It is the state of matter in which electrical resistance of a material goes to Zero abruptlybelow a certain temperature known as critical temperature Tc, which is thecharacteristic of the material. After the discovery of superconductivity in ceramicmaterials by Bednorz & Muller (1986), over past few years, tremendous efforts havebeen devoted to the study of physical properties of high Tc superconductors and theirnon superconducting parent compounds. Numerous experimental studies of thesuperconductors, in both superconducting and normal states, have revealed manyproperties that are substantially different from the characteristic of the conventionalsuperconductors. This in turn raises a serious question regarding whether hightemperature superconductivity derives from the same pairing mechanism known to beresponsible for the conventional superconductors

In order to understand the phenomena of this high temperaturesuperconductivity poly crystalline sample of rare earth oxides and bismuth oxidessuperconductors doped with Pb Nb, Sb or V will be studied by the neutron diffractiontechniques. The objectives are to find the structural and lattice dynamics parameters byelastic and inelastic neutron diffraction techniques. From structural information exactpositions of the various parents and substituted dopants and oxygen atoms will bedetermined. The Rietveld method will be used for the refinement of crystal structureparameters. The structural information will be correlated to zero resistance criticaltemperature which is an important parameter for the high temperature superconductors.

78

REFERENCES

A. U. Ahmed, N. Ahmed, J. Aslam, N.M. Butt, Q. H. Khan & M. A. Atta (1976), J.Poly.Sci. 14,561.

N. Ahmed, M.M. Beg, N.M. Butt, Q. H. Khan & J. Aslam (1977), J. Nucl. Mat. 68,365.

N. Ahmed, N.M. Butt, M.M. Beg, J. Aslam, Q.H. Khan, M.F. Collins (1982), Cand. J.Phys. 60, 1323.

J. Aslam, S. Rolandson, M.M. Beg, N.M. Butt & Q.H. Khan (1976), Phys. Stat. Sol.B77, 693.

J. Bashir, Q.H. Khan & N.M.Butt (1987), Acta Ciyst. A43, 795.J. Bashir, N.M. Butt & Q.KKhan (1988), Acta Cryst. A44, 638.J. G. Bednorz & K. A. Muller(1986), Z. Phys. B64,189.M. M. Beg, J. Aslam, Q. H. Khan, N.M. Butt, S. Rolandson & A. U. Ahmed (1974),

J.Poly.Sci. 12,311.M. M. Beg, J. Aslam, N.M. Butt, Q.H. Khan & S. Rolandson (1974), Acta Cryst.

A30,662.M. M. Beg, J. Aslam, N. Ahmed, Q.H. Khan & N.M.Butt (1979), Phys. stat. sol.

B94, K45.M.M. Beg, J. Aslam, N.M. Butt, Q.H. Khan & S. Rolandson(1974), Acta Cryst.

A30, 662.M. M. Beg, S. Mahmood, N. Ahmed, J. Aslam, Q. H. Khan & N. M. Butt (1981),

Phys. Stat. Sol. B106, K43.M.M. Beg, N.M. Butt, Q.H. Khan & S.U. Cheema (1985), Neutron Scattering in

Nineties, IAEA, Vienna, pp. 539.A. J. Bradlay & A. H. Jay (1932), Proc. Roy. Soc. A136, 210.N.M.Butt, N. Ahmed, M.M. Beg, M. A. Atta, J. Aslam & Q.H. Khan (1976), Acta

Cryst. A32,674.N. M. Butt, K.D. Rouse, & M. W. Thomas(1978), Acta Cryst. A34, 759.N. M. Butt, J. Bashir, B. T. M. Willis,(1988), A44, 396.N. M. Butt, J. Bashir & M. Nasir Khan (1993), Acta Cryst, A49, 171.N. M. Butt, J.Bashir, & M. Nasir Khan (1993), J. Mater. Sei. 28.G. Caglioti, A. Paolleti & F.P. Ricci (1958), Nucl. Instrum. 3, 223.A.K.Cheetham & J.C. Taylor (1974), J. Solid Stat. Chem, 21, 253.D.R. Chipman & A. Paskin (1959), J. Appl. Phys, 30, 1992.A.W. Hewat(1979), Acta Cryst. A35, 248.A.W. Hewat(1985), Chemica Scripta (Proceeding of meeting on advances in powder

diffraction crystallography, Stockholm, May 1985).G. Honjo & W. Watanabe (1958), Nature, 181, 326.R. W. James (1967), The Optical Principles of the Diffraction X-rays, Bell, London.D.W. Jones (1958), J. Poly. Sei, 32, 371.D.W. Jones (I960), J. Poly. Sei, 42, 173.S. Mahmood, N.M,Butt, N. Ahmed, M.M. Beg, Q.H. Khan & J. Aslam (1980), Acta

Cryst. A36,147.J. M. Meese (1983), Neutron Transmutation Doped Silicon, Plenum Press.K. H. Meyer & L. Misch (1937), Helv. Chim. Acta, 20, 232.N. Nilsson (1957), Ark. Fys. 12, 247.H. Rauch(1991), International Atomic Energy Agency Report of consultant'smeeting

on Nuclear Techniques in the development of advanced ceramic technologies,1-4 October, Viena, Austria, pp 4.

H. M. Rietveld(1969), J. Appl. Phys. 2, 65.B. T. M. Willis (1969), Acta Cryst. A25, 227.

79Next page left blank

FRAMEWORK FOR A SUSTAINABLE DEVELOPMENT OFNEUTRON BEAM WORK IN THE SMALLER RESEARCH REACTORS

F.G. CARVALHO, F.M.A. MARGAÇAPhysics Department,ICEN/INETI,Sacavém, Portugal

Abstract

The authors analyse the present situation of research reactors for neutron beam work in the light of thechanges that took place in the nuclear field during the last decades.

Trends in supply and demand of neutron beam time in view of the specific requirements of the techniquesand of the user's community are outlined It is argued that resources, both human and material, should beconsidered in a global perspective, encompassing the national, regional and international levels, where nationalfacilities, mostly low flux research reactors, should be looked upon as a valuable component of acommonwealth of resources to be usefully exploited for the benefit of the neutron user's community at large.The importance of international cooperation to develop a higher level of research reactor utilization isemphasized while suggestions concerning the role of IAEA are made, particularly, to promote the mobility ofscientists and engineers directed from developed to less developed countries (LDC's) where research reactorsare in operation.

The potential of small research reactors in LDC's as an instrument of the country's general scientific andtechnological development is pointed out as well as difficulties commonly experienced and essentialrequirements of a successful performance with emphasis on the importance of establishing close links with thenational scientific community and especially with university groups.

The scientific and technological relevance of neutron scattering techniques is discussed. Reference is madeto the techniques best suited to modest research reactor facilities as well as to the importance of developing alocal competence in instrument design, optimization and construction.

INTRODUCTION

1. At the heart of the problem set proposed to the Advisory Group lie the present andthe future of nuclear reactors for R&D work. Both appear to bear in many waysthe marks of the origin of these complex facilities.Around the world a number of nuclear research reactors of varied designs were builtand have been operated under very diferent conditions and with diferent degrees ofsuccess, during the last 35 years.In Europe (East and West) the number of operable research reactors with a capa-bility for neutron beam work is estimated at 20-30.Nuclear R&D reactors, including MTR reactors, can be classified according to themaximum nominal thermal neutron flux at the core as low, medium or high fluxreactors, the three classes corresponding to less than 5 x 1013, less than 5 x 1014 ormore than 5 x 1014 neutrons per cm2 and second.To give a few examples: the reactors at Delft and Lisbon are in the first category;those at Ris0 and Saclay (LLB) are in the second category; the Grenoble reactor isthe third category. [1] [2].

81

2. The majority of these nuclear reactors were built in the sixties, either out of agenuine necessity to foster the progress of nuclear energy (the case of the industri-alized nations) or as prestige symbols associated with vague aspirations of powerpolitics. The superpowers played a role in this since in some cases they presentednuclear research reactor installations as a gift to countries in their respective areasof influence.Reactors specially designed as radiation sources for non-energy applications, par-ticularly for neutron beam work, generally came at a later time and are few innumber.Profound changes in the situation of the nuclear power industry have occurred. Theeffects of these changes together with those of the economic recession of recent yearsbear upon the nuclear ou nuclear-related scientific and technical community.On the other hand the progress in methods and instruments, particularly, radio-metric techniques using neutron beams, and the slow but progressive spreadingamong researchers of the awareness of their specific advantages and of expertise intheir utilisation, apparently leads to a steady growth of user demand for access toneutron beam facilities, notably in Europe.We are thus facing a situation where several existing sources are reaching the end oftheir useful life (for physical or political reasons); where pressing problems are posedby the reprocessing and storage of spent fuel where "expertise (in nuclear engineer-ing) in human terms is a declining resource" [1], affecting both design and operationcapabilities; while the neutron user's community is probably growing more rapidlythan at any time in the past.

3. The user's group in both the developed and the less developed countries, is com-posed largely of researchers who wish to carry out basic research, mostly orientedand "far from the market". A demand for services in very special areas also ex-ists.Two examples are the aerospace industry (neutron imaging) and health care(neutron capture therapy).Quite often neutron techniques are used alongside with other techniques to charac-terize specific sample and material properties. This means that it is seldom pos-sible or practical to address a scientific problem using a single experimental tech-nique and that the availability of neutron radiometric techniques will not by itselfsatisfy the requirements of a research project. Cooperation between research groups,normally requiring a certain degree of human mobility, can be the solution to thistype of technical limitations. Also, the evaluation of neutron data can be a time-consuming task, this being another reason for the user of neutron techniques todivide his or her time between different work places.

82

The development of the application of neutron beam techniques cannot be separatedfrom the development of the techniques themselves. This is a kind of work that re-quires the actual or potential availibility of neutron beams at the researcher's homeinstitution — whatever that means, given the current practice of highly "flexible"employment.

4. The situation of demand and supply that has been outlined, in view of the specificrequirements of the techniques and of the user's community, recommends thatexisting resources (both human and material) and resource development be consi-dered in a global perspective, encompassing the national, regional and internationallevels. On the one hand, and anticipating a major shortfall of neutron sources at thedawn of the 21st century, imediate efforts should be made to ensure that a limitednumber (5 - 6) of specially designed state-of-the art high flux neutron sources (ofthe reactor and/or the accelerator type) would be available world-wide around theyear 2000. Access to these facilities should be open on a fair basis to the inter-national scientific community, through adequate schemes of cooperation — an endthat should be easier to attain in the present post-cold war world than in the past.On the other hand, measures should be taken to guarantee a more efficient use ofexisting facilities.In the case of medium and high flux reactors in the scientific and technically moreadvanced countries it would seem unreasonable to increase reactor power or to imple-ment other modifications that may require new safety assessment procedures. Oneshould preferably concentrate on improving the quality and increasing the num-ber of neutron beam-tube facilities, including new and upgraded instrumentationand optimally designed spectroscopy equipment. An increase of available beam--time and a reduction of spare capacity should be the result. Such a course of ac-tion would require, of course, a certain amount of capital investment and more fundsfor current operation expenses.

5. Also, national facilities in the less developed countries — mostly to be classified inthe category of low flux research reactors — should be looked upon as a valuablecomponent of a commonwealth of resources to be usefully exploited for the benefitof the neutron user's community at large.This would represent a major step towards improving the utilisation of beam portfacilities which in several cases have been there for many years without adequateuse or no use at all as a consequence of the lack of trained personnel and lack offunds but also due to organizational problems and insufficient motivation at variouslevels - political and scientific.

83

From our own experience we know well the catalytic effect of external inputs, ma-terial and also imaterial, on the availability of funds and general support for thedefinition and execution of programmes and projects in land. They can be used asarguments to satisfy political decision makers and they do work more often than not.From this point of view, but also from the point of view of the effective momentumthey can communicate to projects, the forms of the visiting scientist and of thescientific mission are particularly relevant forms of cooperation.If national facilities are perceived as an integral part of an international whole ofresources that are complementary in the pursuit of common goals — namely theadaptation of supply and demand for the utilization of neutron beams — that per-ception will constitute an incentive to cooperate. The question is to what extent isthis actually true.We believe it is true and that the importance of local facilities (as opposed to cen-tralised large instalations) is being recognized under the pressure of circumstances:"(...) apart from their training capabilities, they provide neutron beams for instru-ment development, monochromator and detector improvements and the testing ofcomponents. In addition they allow good science and technology to be performedfor those areas where the highest fluxes are not required" [1].

RESEARCH REACTORS IN DEVELOPING COUNTRIES

1. The role of a small research reactor in a developing country deserves some thought.In the first place it is not an easy task to define a developing country. For thepresent purpose a good indicator is probably the percentage of GNP that is spentin R&D activities.If we place the limit at 1% then we will have included Portugal, Spain and Greecein the group of developing countries.It is probably true to say that the decision to built a research reactor in a developingcountry was taken at the time for a combination of reasons: aspiration for a nuclearpower programme, more or less vague military ambition, prestige. At the politicallevel the urge to develop general purpose scientific and technological capabilities hasprobably played a minor role in that decision.Typically, as time went by, the original intentions were largely frustrated. Also,since it is generally easier to make decisions to invest capital than to finance currentexpenses the facilities lived a stagnant life, until, eventually the original justifica-tions having disappeared, a fight for survival set in.Notwithstanding this long and sometimes despairing evolution, one should contendthat there is a real value in research reactors as tools for carrying out fundamen-tal and applied research work in various disciplines and for the training of young

84

scientists and technicians in basic science and modern technology. In fact there isno hope for a developing country to assert itself in the modern world without a hugeeffort to raise the level of scientific and technical education and professional trainingof the younger generation. The development of R&D activities is an indispensablecomponent of that effort. Also expertise in nuclear techniques has a wide horizontalrange of applications in almost all spheres of social activity.A successful activity in science and technology requires particularly demandingforms of organization and management of resources (human, material and finan-

cial) that are often beyond common standards in developing countries. This is themore so in the case of a nuclear reactor installation for reasons of complexity, safetyrequirements and sheer dimension — a research reactor has a good chance of beingby far the largest scientific and technical single facility in a developing country.

2. In order to fulfil a useful function a research reactor should establish as many linksas possible with the scientific and teaching community of the country and also withthe international scientific community. Isolation at home should be fought withperseverance. This can be done especially by establishing contacts with universitygroups that are potential users of neutron techniques and by introducing this tech-niques to interested colleagues.There should be a disposition of the reactor group (or the neutron scattering group)to offer experiment design capabilities and support services to carry out experimentsand to analyse data.To attract teachers and students of different levels some thought should be given tothe preparation of demonstrative experiments.To perform these functions a small motivated team (4-5 people) of good scien-tific and technical level is necessary. This means permanent posts with good careerprospects for the small group of scientists and engineers. The group should be in-volved in a few scientific projects of their own, that should include some form ofpartnership or cooperation with foreign groups.A significant part of the group's activity should be dedicated to the development,construction and improvement of instrumentation. This is important for the group'swork and as a source of technical spin-offs to the community.The group would also provide the necessary technical and scientific basis tosupport forms of international cooperation to be carried out "in situ". The formsthat appear to be the most interesting are the following:

• visits of scientists and/or engineers to help design, build and install new or dis-placed instruments, with the right to become regular users of the instruments.Displaced instruments may be instruments from facilities that have been shut

85

down or also instruments from larger facilities (to be adapted locally or refur-bished) for which new instruments have been substituted in their original work-ing place.

• scientific visitors interested in having access to existing instruments or beamtube facilities where beam time is almost entirely free, to carry out "odd" ex-periments, to perform preliminary analysis of special samples and or for otherpurposes.In certain cases, the lucky visitor would be suprised to find locally exceptionallygood people at the level of technicians and also bright young researchers thatare unfortunately not adequately supported in their home institutions.

INSTRUMENT DESIGN

1. The importance of the development and application of new and advanced mate-rials is well understood in the developed and in most developing countries. It isalso common knowledge in materials science that relating bulk properties to micro-scopic structure is an essential step towards the development of new materials andnew processing technologies. Namely, information is required on relations betweenstructure, properties and performance and how they are affected by processing.Neutron scattering techniques have proved very fruitful in the investigation of struc-tural properties of complex materials. As a tool for the characterization of mate-rials, thermal neutrons offer several well-known and unique advantages over othercommon probes such as protons, electrons or X—Rays. Since they carry no elec-tric charge, they are deeply penetrating. Because of their low energy they arenon-destructive. Low energy also makes them ideal for inelastic scattering stu-dies. However, inelastic scattering cross-sections are invariably very low and suchstudies are best left to high flux reactor facilities. On the contrary, in elasticscattering experiments a substantial fraction of the incident neutron flux is actuallymeasured, and these studies can be successfully carried out at reactors with modestfluxes. Furthermore-and this is a key point-some of the neutron elastic scatteringtechniques are particularly well suited when it comes to tackle a few importantproblems of technological relevance on a level likely to lead to practical progress.Generally the techniques which are felt to be particularly well suited to modestresearch reactor facilities [3, 4] and which probe the microscopic structure of mate-rials, fall into two broad categories depending upon the magnitude of the angle 6 bywhich the neutron beam is scattered:

86

• Small Angle Neutron Scattering (SANS) (0.01< 6 < 10°) which probes homo-geneity at a scale of 10-1000 A, such as pores or precipitates in a matrix, macro-molecules in solution, etc.

• Powder Diffraction, for larger angles (0 > 10°) which explores structure on theinteratomic length scale (1-10 A).

• A third interesting technique is neutron reflectrometry for surface studies.

The techniques have different degrees of complexity and different budget require-

ments. They can be implemented at reactors with ambient sources and demand aminimum expertise and design capabilities in the fields of fine mechanics, radiationshielding, electronics and data accumulation and processing.We have some experience in the implementation of these techniques. Our first in-strument was a time-of-flight diffractometer for powder diffraction and the second,now under construction, is a SANS facility. In both cases we have learned that de-sign optimization is the key to get the most out of the available reactor flux, throughthe instrument.

2. A neutron scattering facility is often thougth of as a standard black box (which canbe installed in no matter what type of neutron source): a neutron beam goes inand data comes out. Such a facility when installed at a low or medium flux reactorwould yield a count rate lower by one or two orders of magnitude, respectively, thanthe corresponding count rate when installed at a high flux reactor.On this basis it is sometimes argued that it is pointless to try and install moststate-of-the art scattering facilities at the smaller research reactors with ambientsources.We have a different view. To start with, we reject the concept of the standard blackbox facility. We argue that each case is a case, meaning that a specific lay-outhas to be studied and an optimized design carried out for each particular facility, tomake the most efficient use of the available neutron source and minimize all possiblesources of loss. That is, for each instument a particular design can be developed,optimized for the instrument's required performance and for the characteristics ofthe reactor installation. This has been demonstrated for different types of neutronscattering instruments. In our case we have studied the optimization of a diffrac-tometer for stress measurements with the result that a count rate gain of one orderof magnitude could be achieved relative to the use of a standard diffractometer forstress measurements [5]. Taking into account that the instrument design should beoptimized for the actual installation characteristics, it is no longer true that theratio of the count rates of two such instruments installed at two different reactors

87

is the ratio of the corresponding maximum nominal thermal neutron fluxes (just asthe ratio of thermal neutron fluxes is not the ratio of the reactor thermal powers).For the last four years, we have been studying the optimization of SANS instru-ments for different types of neutron sources (pulsed, steady-state); different colli-mation assemblies (guide tubes, single pin-hole and multiple pin-hole collimators),and different detector assemblies and sizes. The main results have been reportedand discussed [6—11].

One result is that the design which had been considered for about 20 years to bethe optimum one - the equal-flight-path design - does not lead to the highest countrate for a given measurement with fixed resolution. Its main handicap is the use ofan effective neutron source area which is invariably a fraction of the total availablearea. The design requires a source area equal to twice that of the detector countingcell (around 2 cm2) whereas the available source area at a beam port is about oneorder of magnitude larger.We have shown that when the full source area available is used the count rate gainrelative to the conventional design can go up by a factor equal in the limit to 4 or16, for single and multiple pin-hole collimation, respectively. The actual gain factordepends on the physical constraints and the characteristics of the installation siteas well as on the actual size of the source area.Our SANS instrument optimization results have been experimentally confirmed atRis0 National Laboratory [12]. We recognize that the implementation of a largesource area design at a medium or high flux reactor is often very difficult becausea large number of scattering instruments are usually crowded around the neutronsource. This is not the case, in general, for the lower flux reactors. Wherever pos-sible the use of the large source area design with multiple aperture collimation isrecommend. Our neutron scattering group will be happy to collaborate in the de-sign of SANS instruments tailored to match specific neutron source characteristics,space constraints and other installation room characteristics.

FINAL REMARKS

To develop a higher level of research reactor utilisation, particularly for neutronbeam work, international cooperation is a decisive factor.

Increased mobility of scientists and engineers from developed to developing countries— that is, contrary to the normal trend — is essential. Also, existing capabilities in theless developed countries should not be underestimated. They can contribute beneficiallynot only to specific research activities in the developed countries but can also be usedto enhance capabilities in other developing countries. It is probably true that up to acertain degree of complexity problems experienced by groups in certain less advanced

countries can be better solved — more economically and more efficiently — by groupsworking in other developing countries that have managed to attain a higher level ofscientific and technical expertise.

Solved and unsolved problems in different developing countries often do not coincide.To take advantage of this an effort should be made to foster communication and diffusionof technical information between research groups and institutions in different developing

countries.The availability of good auxiliary services and infrastructures, mainly workshops, is

an important asset. Related to this and equally important is expertise in design andconstruction of instruments, special parts and components.

The Agency could act as a vehicle to spread information on this capabilities, andeventually act as a match-maker and sponsor of contracts between centers in developingcountries for the supply of certain specific types of equipment and services.

REFERENCES

[1] "Study Panel on Neutron Beam Sources", Advisory Comm. on the Large Installa-tions Plan, CEC DGXII, Draft Report Oct. 199Q.

[2] "Nuclear Research Reactors in the World", Reference Data Series ne 3, IAEA,Vienna, 1991.

[3] Springer, T., in IAEA Advisory Meeting on the "Use of Research Reactors in BasicResearch in Developing Countries", Sacavém, Portugal, 1983.

[4] Axe, J.D., in IAEA Report of Consultant's Meeting "Nuclear Techniques in theDevelopment of Advanced Ceramics Technologies", Vienna 1991, p.9.

[5] Margaça, F.M.A., in "Measurement of Residual and Applied Stress Using NeutronDiffraction", Ed. M.T. Hutchings and A.D. Krawitz, NATO ASI Series E, Vol. 216,Kluwer Academic Publishers, 1992, pp.301-311.

[6] Margaça, F.M.A., Falcâo, A.N., Salgado, J.F. and Carvalho F.G., Physics B 156&157(1989) 608-610.

[7] Margaça, F.M.A., Falcâo, A.N., Salgado, J.F. and Carvalho F.G., Nucl. Instr. andMeth. A274 (1989) 606-607.

[8] Margaça, F.M.A., Falcâo, A.N., Sequeira, A.D., and J.Salgado, J.F., J. Appl. Cryst.24 (1991) 531-536.

[9] Margaça, F.M.A., Falcâo, A.N., Salgado, J.F. and Carvalho, F.G., J. Appl. Cryst.24 (1991) 994-998.

89

[10] Margaça, F.M.A. and Salgado, J.F., Physica B 180&181 (1992) 947-950.

[11] Margaça, F.M.A. and Salgado, J.F., "Characteristics and Potentialities of the Asym-metrical Design for SANS Instruments" (1993) to be published.

[12] Falcäo, A.N., Pederson, J.S. and Mortensen K., "Optimum Intensity in Small AngleNeutron Scattering. An Experimental Comparison between the Symmetrical andAssymetrical Set-ups", (1992) to be published.

90

BASIC RESEARCH IN SOLID STATE PHYSICSUSING A SMALL RESEARCH REACTOR

V. DIMICInstitut Jozef Stefan,Ljubljana, Slovenia

Abstract

The 250 kw TRIGA Mark n reactor of th<s J. Stefan Institute In I jxibljana, Slovenia, is alight water reactor which is used for basic research in the following main fields: solid statephysics, neutron radiography and autoradiography, neutron dosimetry, reaktor physics,neutron activation analyses, etc. In the field of solid state physics the rotating crystaltime-of-flight spectrometer together with the cold neutron source using solid methane as amoderator has been used to measure incoherent inelastic neutron scattering spectra inorder to study optical and acoustic vibrations, phase transitions, and diffusive motions indifferent materials. Nondestructive studies using neutrons are performed for inspection ofinternal distribution of elements in alloys, structure studies in polycrystalline materials, etc.For these studies the following methods are used: neutron radiography with thermal,resonance and fast neutrons, neutron induced autoradiography using solid state nucleartrack detectors and radiography with back-diffused electrons-

Research reactors with the flux in the region of about 1013 n/cm2 s may serve as animportant research tool which is in many fields competitive with the reactors in the higherflux region. Use of low (or medium) flux reactors for research has often demanded a higherdegree of ingenuity and a sense of pioneering, whic are important characteristics forscientific research of high quality. In order to achieve this, the choice of the reactorinstrumentation has to be implemented which makes optimum use of this flux, the reactorneeds a certain amount of auxiliary "background laboratories" and the link between anuclear centre and a University. The acquisition and installation of nuclear reactor hasmainly been justified on the following ground: to produce isotopes, to support basic (andapplied) research in various fields and as a tool for training in the nuclear sciences. Theseobjectives have been met with different degrees of success. As an example of how theseobjectives could be fulfilled rather successfully with a small research reactor like a 250 kWTRIGA Reactor the basic research programme in solid state physics carried out at the J.Stefan Institute in Ljubljana (Slovenia) will be discussed in detail.

1. TRIGA Mark U Research Reactor at the J. Stefan InstituteThe 250 kW TRIGA Mark II reactor began operation in May 1966. It is a light water

reactor, with solid fuel elements in which zirconium hydride moderator is homogeneouslydistributed between 20% or 70% enriched uranium. The maximum neutron flux in thecentral thimble is 1.1013 n/cm2 s. A 40 position rotary specimen rack around the fuelelements, a pneumatic transfer rabbit system, as well as a central thimble are used forirradiation of samples. The thermal neutron flux at the rotary specimen rack is 1.4 x 1012

n/cm2 s.

91

Other experimental facilities include two radial and two tangential beam tubes, agraphite thermal column and a thermalizing column.

2. Research in solid state physics by neutron scatteringThrough neutron inelastic scattering experiments detailed information can be gained

on phonon spectra and dispersion curves, phonon lifetimes, thermal diffusion of atoms, thetime dependence of the case when a sample contains atoms with a high incoherent orcoherent cross section for neutrons. For instance, neutron scattering by hydrogenoussubstances is almost entirely incoherent due to the large incoherent cross section of theproton. Therefore, when hydrogen atoms are present in the molecule, neutrons arescattered mainly by these atoms, providing relatively easy interpretation.

2.1. Rotating crystal time-of-flight spectrometer

Incoherent inelastic neutron scattering spectra can be measured by cold neutrons(energy gain method) with a neutron energy lower than 5 meV (a wavelength is higher than4A. The spectra of scattered neutrons are very often measured by a time-of-flight method.A pulsed beam of almost monoenergetic neutrons, produced by rotating single crystals oflead or copper, is scattered by a sample into a bank of neutron detectors arranged atdifferent scattering angles to the incident beam direction at a distance 2- 3 m from thesample. The energy spectrum of the scattered neutrons is analysed by measuring thescattered neutron time-of-flight from the velocity selector to the neutron detector bymeans of a multichannel analyser.

Such a spectrometer, where the neutron spectra are measured at four scattering anglessimultaneously with 4 banks of He-3 detectors, was built at our reactor. A cold neutronsource using solid methane as a moderator was placed in the tangential beam hole. Withthis cold neutron facility it is possible to perform experiments which otherwise could noteasily be carried out with a low flux reactor^1"3).

The rotating crystal time-of-flight spectrometer together with the cold neutron sourcerepresent very valuable experimental tool for research in the field of solid state physics byneutron inelastic and elastic scattering. At our reactor this method is used for the study ofoptical and acoustic lattice vibrations, phase transitions, and diffusive motions in

— ferroelectrics and antiferroelectrics:KH2PO4, CO2Sr(CH3-COO)6, PbCa2(CH3-COO)6,

— liquid crystals: MBBA, PAA, Na-palmitate, Na-stearate, anisalazine, etc.

— biological samples: LiDNA, NaDNA

— hydratization of cement.

23. Liquid methane cold neutron moderator

Thermal and subthermal neutrons are a unique tool for research on condensed matterphenomena because their wavelengths are comparable to those of bound atoms. The

92

magnitude of the energy transfer involved in these phenomena is about 0.1 to 100 meV. Inorder to detect these small energy changes one is often forced to use very slow neutrons;4 A to 6 A is a convenient wavelength region. Reactors used as slow neutron sources inthese scattering experiments usually have moderator temperatures near 60°C. Thus, aMaxwellian spectrum of neutrons emerging from a reactor at these temperatures has amaximum around 1-2 A and the flux falls off rapidly towards higher and lower energies.Therefore, cold neutrons represent only about 1% of the total spectrum. It is evident thanan increase in the flux might be achieved if the reactor moderator temperature could belowered. For neutrons of wavelength 4 A, an increase of an order of magnitude is possible,and much higher increases are obtained at higher neutron wavelengths.

Many materials have been used as moderators, for example, liquid hydrogen, severalliquid hydrogen/deuterium mixtures, liquid deuterium or heavy water, methane and methylalcohol. Liquid hydrogen and methane have the best moderating properties because oftheir small-down length, long scattering cross section and because atoms are loosely bound,which gives a high density of states at low energies.

At our reactor, TRIGA Mark II, a time of flight spectrometer has been installed whichis used for cold neutron scattering experiments. The flux of neutrons in the low energy tailof the thermal spectrum for the TRIGA reactor was inconveniently low. It was thereforedecided that a cold source should be placed in the tangential beam hole in order toincrease the cold neutron flux by a factor of five to six.

The loop was installed in the reactor and at first it was tested with methyl alcohol as themoderator; however, a large decomposition rate of methyl alcohol was found. Methylalcohol was replaced by methane, with a boiling point at -161.5°C and freezing point at-148°C.

to ballast reservoir. liquid N2

reactor tank

CORE shielding Bi filterliquid N2

. - . f • -=^ in ,Hs.-f,/Mn 2c/£&f& ^̂ ^̂——. -.-- I KWM'tm ws/7Jm ^̂ ^̂

moderatorchamber

Fig. 1. General layout of cold neutron source facility.

93

The neutron cold source was inserted into the tangential beam tube 15.0 cm indiameter, traversing the graphite reflector (Fig. 1). The time of flight spectrometer isplaced on one side of the beam tube and the service end of the cold source is on the otherside. Vacuum pumps, inlet and outlet pipes for methane and liquid nitrogen, and thecontrol system are situated at the service end of the cold source. The berillium filter withcollimators is placed in another vacuum jacket. With this arrangement, it is possible to takeout from the beam hole either the beryllium filter or the cold source. The vacuum ismaintained by a diffusion pump and is better than 10~4 torr.

The optimum shape and position of a moderator chamber for our cold source has beencarefully investigated^1). We have found that the optimum moderator chamber should beshaped in a form of a black-body scatterer.

The spectra of neutrons, slowed down in the black body shaped moderator chamberusing methyl alcohol and methane as moderators, were compared with the spectrum ofcold neutrons produced by a graphite slab with a length of 15 cm and diameter of 13.8 cmand located at the .centre of the reactor core. The results are given in Table 1. where themoderator efficiencies are tabulated.

A measurement taken with the moderator chamber filled with water at ambienttemperature has been used as a reference point.

Table 1. Relative cold neutron output

WaterT = 300 K

1.0

GraphiteT = 300 K

0.62

Methylalcohol

T = 300 K0.93

GraphiteT = 100 K

0.66

Methylalcohol

T = 88 K4.3

MetlaneT = 88 K

5.1

We conclude that the construction and operation of a cold neutron source is a fairlystraightforward matter and that, with a source such as that described, reliable andcontinuous operation can be readily achieved. During the operation normally lasting 100 h,no special checking is required. However, the gain of cold neutrons is quite considerable.With such a cold neutron facility it is possible to perform experiments which otherwisecoud not easily be carried out.

3. Nondestructive studies using neutronsIn the last thirty years significant efforts have been devoted to developing different

nondestructive methods using neutrons from the research reactors for inspection ofinternal distribution of elements in alloys, structure studies in polycrystalline materials andlong range fluctuations of molecular or magnetic density in matter. The most usefulmethods in order to investigate all this phenomena are neutron radiography especiallymicroneutronographic and neutron induced autoradiographic techniques, neutrondiffraction and small angle scattering of neutrons. At the reactor in Ljubljana research inthe fields of neutron radiography stimulated the development of related fields likemicroneutronography, neutron induced autoradiography and radiography withback-diffused electrons^15'23).

94

3.1. Neutron radiography

Neutron radiography is a valuable non-destructive method, which can be placed alongside the more familiar and well- developed X-ray and y-ray radiography. However, neutronradiography should not be regarded as a method competitive with X and y-ray radiography,but has to be considered as an additional radiographie method, which providescomplementary and, in some cases, unique information relative to other radiographiemethods.

In general, there are two ways in which neutron radiography can be applied:

a. As a high resolution neutron radiographie method, capable of investigating in aqualitative and quantitative manner the microstructures and composition of a specimen.This technique is called microneutronography.

b. As a defectoscopic method capable of detecting on a macroscopic scale variousdetrimental defects in the objects being investigated.

3.1.1. Principles of microneutronography

Microneutronography is a high-resolution radiographie method for inspection of theinternal distribution of highly neutron- attenuating elements in alloys. The method can beused either for viewing the internal microstructure in the sample or could possibly beapplied in a quantitative manner for the determination of the internal composition of aspecimen by means of microdensitometric readings. In microneutronography applied toviewing the internal microstructure an image is obtained by placing a specimen, usuallythin (from about 100 fim and up to a few 1000 ftm), in intimate contact with a suitableneutron image detector and exposing it to a collimated beam of neutrons. The primaryimage represents the spatial variations in the intensity of the transmitted neutron beamowing to different attenuation by the various phase in the specimen. Thus the chemicalheterogeneity is revealed as a variation in photographic density. The microneutronographis then obtained by enlarging the primary image by means of the light microscop. Theresolution of the neutron-imaging technique must be adequate to allow the observations ofinternal microhetereogeneities at useful magnification of the primary neutron radiographfrom at least 5 x up to 100 x. A resolution of about 5 pm can be achieved by usingtrack-etching techniques together with boron-10 converters.

The aim of quantitative microneutronography is to determine the concentrations ofstrong neutron absorbing elements. Quantitative microneutronography can be appliedeither to obtain concentrations averaged over structural details in the area of the scanningwindow of the densitometer and throughout the specimen thickness, or concentration in aparticular phase.

3.1.2. Neutron induced autoradiography

Microneutronography is a suitable method for the qualitative and quantitativeinvestigations of the distribution of strong neutron-absorbing elements, especially of lightelements (B, Li). However, the method is limited by its sensitivity and spacial resolution.The complementary method to microneutronography, with required sensitivity and spacial

95

resolution is neutron-induced autoradiography. In cases where a specimen containselements like B, Li or U, which by themselves emit charged particles upon neutronabsorption, the presence of these elements can be revealed without any intermediateconverter screen directly by detection of reaction particles with a suitable detector. Itshould be noted that in this case the collimation of neutrons is no more required. In thismethod the specimen to be examined is polished and appropriate detector is placed againstthis surface. The autoradiograph represents the distribution of particular elements on thesurface in contact with the detector. As a detector of charged particles could be usedvarious Solid State Nuclear Track Detectors (SSNTD) where the track-etching is used.

4. ConclusionsThis paper has briefly summarised the variety of experimental activities in solid state

physics using a small (or medium) research reactor and which can easily be a subject offruithful cooperation between laboratories in different countries. The research activities inthis field have generated development of a variety of experimental facilities and otheraccessories to provide a self-sustaining growth in science. As to the choice of an eventualbasic research programme one should try to consider the infrastructure available in thecountry and the connections to the already existing activities in the country.

References

1. V. Dimic, J. PetkovSek, Nucl. Instr. Meth., 96, 385 (1971).2. V. Dimic, J. PetkovSek, Journal of Scient. Instr., 4, 905 (1971)3. V. Dimic, Nucl. Instr. Meth., 107, 405 (1973).4. V. Dimic, M. Osredkar, Phys. Lett. 25 A.123 (1967).5. R. Blinc, V. Dimic, Phys.Lett., 31 A,10, 531 (1970).6. R. Blinc, V. Dimic, J. Pir§, Mol. Cryst. and Liq. Cryst., 14,97 (1971).7. V. Dimic, L. Barbie, R. Blinc, Phys. Stat. Solidi (b), 54, 121 (1972).8. V. Dimic, M. Osredkar, Mol. Cryst. and Liq. Cryst., 19, 189 (1973).9. V. Dimic, M. Osredkar, J. Slak, Phys. Stat. Solidi (b) 59, 471 (1973).10. V. Dimic, M. Osredkar, Phys. Stat. Solidi (a), 30 (1975).11. B. Cvikl, M. CopiC, V. Dimic, Jour, de Phys. 36,107 (1975).12. U. Dahlborg, V. Dimic, A. Ruprecht, Phys. Scripta Scand., 22, 179 (1980).13. U. Dahlborg, V. Dimic, B. Cvikl, Phys. Scripta Scand., 37, 93 (1988).14. B. Cvikl et al., Liquid Crystals,.? (5) (1990).15. M. CopiC et al., Materialprüfung 18,171 (1976).16. R. Ilié et al., Solid State Track Detectors, Pergamon Press Oxford, 655 (1980).17. R. Ilié et al., Solid State Track Detectors, Pergamon Press Oxford, Suppl. 3,183 (1982).18. J. Rant, R. Hie, Atomic Energy Review 15, 327 (1977).19. R. Ilié, J. Rant, F. Sirca, Proc. Conf. Radiography with Neutrons, Brit. Nucl. Energ. Soc.,

London, 139 (1975).20. R. Hie, A. Podgornik, M. NajZer, Proc. Int. Conf. for the Study of Bauxities, ALumina and

Aluminium, Athens, Vol. 3, 126 (1978).21. J. Rant, G. Pregl, Nucl. Instr. Meth., 150, 317 (1978).22. J. Rant et al., Nukleonika, 25, 1261 (1980).23. J. Rant, Nucl. Tracks and Radiation Measurements, Pergamon Press (1990).

96

THE NEW SWISS SPALLATION NEUTRON SOURCE SENQAND ITS PLANNED DAY-ONE INSTRUMENTSStatus as of spring 1993

O.S. BAUERPaul Scherrer Institute,Villigen, Switzerland

Abstract

The paper briefly reviews the concept and layout of a cw spallation neutron source forcondensed matter studies, SINQ, under construction in Switzerland. Without involvingfissile or fissionable material this technique allows to build a medium flux neutron sourceessentially with available technology. A target development program to make optimumuse of the proton beam driving the facility, is described and performance estimates for thesource are given. A carefully designed experiment support system inlcuding a coldmoderator, super mirror coated neutron guides and beam shutters incorporated in thebulk shield will help to optimize the performance of the instruments. Six instruments underconstruction and planned for operation when the source will be commissioned aredescribed briefly.

1. Introduction

This paper was prepared upon request by the organizers as one of the introductory talks at theIAEA-advisory group meeting on Uses of Research Reactors for Solid State Studies held atVienna from March 28 to April 1, 1993 in order to outline current activities in a smallindustrialised country in the field. It is intended on the one hand to spread information on newdevelopments, in particular on the implications of using spallation as an alternative method forneutron release from nuclei and on the other hand to serve as a starting point to discuss basicinstrumentation needs at a low-to medium flux neutron source in order to implement a rathercomprehensive research programme using neutron beams.

2. The incentive for a new neutron source in Switzerland

Since 1957 Switzerland has been operating and utilizing its light water pool reactor SAPHIRfor irradiation and beam hole based research. After two upgrades which included the provisionof beryllium reflector elements at the core position adjacent to the neutron beam tubes, thereactor was operated at a power level of 10 MW for several years. The effect of the use ofberyllium elements between the fuel and the beam tubes was an increase in useful thermal fluxand a decrease of fast neutron (and ^-radiation) in the beam ports. While not eliminating totallythe need for filters in the beam, their thickness could be reduced. In this way the Berylliumelements are of double benefit to the neutron scattering instruments. Fig. 1 shows the coreconfiguration of the reactor and the instruments arranged at the different beam tubes. It isremarkable that triple axis spectrometers have been used rather successfully on this reactor.

Based on these available instruments a neutron scattering tradition has developed whichresulted in a successful research programme as documented by the large number of guestscientists and publications per year and the annual reports published regularly by theLaboratory for Neutron Scattering (LNS).

97

I—o 3m

Diffractometer withmulti-counter (DMC) Two axis

spectrometer

Triple axisspectrometer

Triple axis spectrometer

!•) Reinforcement rod

SO] Fission chamberH Beryllium elementffl Standard fuel elementH Control absorber element

<-Tangential tube(unused)

Beam tubeused forradiography

Figure 1: Core disposition and arrangement of neutron spectrometers at the P SI-reactorSAPHIR

98

In 1992 a review of the reactor safety under the rules now in effect suggested that theoperational safety margin might not be sufficiently large at 10 MW and the reactor power wasreduced to under 7 MW.

Independent of this setback it had been decided in 1987 that Switzerland was going to maintainand expand its neutron-based research activities on a longer term basis and to provide a newneutron source for its national and international community of neutron users. Advantage wasto be taken of the planned upgrade of the isochronous proton cyclotron at the then SIN (SwissInstitute of Nuclear Research, now part of PSI). Since this accelerator had been developed forpion and muon production only, about 80% of the accelerated beam used to end up in a beamdump. The cyclotron was orginally designed for operation at 590 MeV, 100 uA.

Due to the very low loss level in the machine it was concluded that an upgrade to the regimearound 1.5 mA was possible with a new 72 MeV injector, working on the same principle as themain ring. In order to make operation at this current level a real option, most of the beamtransport line and the meson production targets had to be rebuilt with extra shielding and moreintense cooling especially in those positions, where beam losses would be unavoidable. Thistask was completed in 1991. Since this time the upgrade of the ring accelerator itself is inprogress, with a planned yearly increase in beam current by some 250 uA. Currently 500 fiAcan be delivered. The goal for 1993 is 750 uA. When operating at a higher current level, theusage of the accelerator could be expanded by adding a suitably designed target to create amedium flux neutron source. Under the given circumstances this would be a relatively cheapway to ensure the future availability of neutrons for the country. At the same time it seemedpossible to meet the growing demand for long wavelength neutrons by incorporating a coldmoderator in the design from the very beginning.

3. Brief review of the source concept

Neutron scatteres were involved in the decisions made early on in the conceptual planningstage and existing experience from other neutron sources was incorporated to the largestpossible extent, subject of course to existing boundary conditions.

One of these boundary condition was that the PSI-cyclotron is operating at an rf-frequency of51 MHz but apart from this has no macro pulse structure. Therefore, the concept of SINQ isthat of a steady state neutron source and hence aims at providing the highest possible neutronflux that can be generated from the available beam power and at offering the maximumpossible space for instruments. For this reason the proton beam is injected into the target fromunderneath, making 360 degrees around the target accessible for beam extraction. The conceptprovides for a multiple containment of all potentially volatile radioactivity produced inside thetarget block itself. For this purpose a double-walled steel tank with contained atmospheres willsurround the target and the moderator tank. A perspective view of the facilities is shown inFigure 2. The target block, which is shown partly cut open to display the space in which thesteel tank was inserted, will surround a 2 m diameter heavy water tank in whose axis the targetwith essentially cylindrical geometry is positioned (Fig. 3). The D2O will guarantee a long lifetime of thermal neutrons and hence a high thermal flux. Two spectrum shifters (cold or hotmoderators) can be located near the maximum of the thermal neutron flux in the D2O-tank.One of them will be a 20 litre liquid D2 moderator, feeding neutron guides and one pair ofbeam tubes. A bundle of 7 neutron guides will serve to transport cold neutrons to the neutronguide hall, where the majority of instruments will be located.

All beam tubes are designed as twin pairs to minimize the flux pertubation in the D2O and stillhave a large number of possible instrument positions around the target block. More details can

99

be found e.g. in the status report delivered at ICANS XI [1]. The whole of the servicesbuilding shown behind the target block will be located inside the target hall. Its construction isdue to start early in April, 1993.

main targetshield

target services andcontrols building neutron

guide hall

used targetrepository

proton beamtransport line

Figure 2: Partial cut-away view of the SINQ facility located in the neutron target hall. Theproton beam line partially shown in the foreground allows injection into the target fromunderneath. The target shielding is shown in partially cut open to display the space where theatmospheric containment was inserted and the penetrations for the neutron beam ports andcold source insertion. Adjucent to the target shielding and located above the neutron guideshielding cavern is the target services building, completely located inside the target hall.Seven neutron guides, two of which are visible, lead into the neutron guide hall.

4. Target development programme

Together with the maximum neutron yield that can be obtained, safe and reliable operation of atarget at the anticipated beam power level of nearly 1 MW is a prime concern in SINQ. Theproton current density will be 25 |iA/cm2 under normal conditions and can go substantiallyhigher for short times, if the upstream meson target fails or the beam is incidentally steeredpast it. Inititally we inteded to start operation with a liquid lead-bismuth target which has anumber of attractive features. However, in view of the lack of experience world-wide withsuch targets, we decided on an approach which puts maximum emphasis on operational safetyand hence availability of the source [2].

100

C\~

Shielding and heattransport zone

Double containmentank

Cooled inner ironf shield

Cold D2-moderator

Neutron producingzone

Cooled beamtube shielding

Thermal neutronbeam tube

Moderator tankcentral tube

D2O moderator

H2Û layer

Proton beamcollimator

Proton beammonitor

Figure 3: The inner parts of the SINQ target block, showing the target and its surrounding£>20 moderator tank which are contained within two concentric steel tank structures withcontrolled atmospheres

101

Starting with a relatively simple design of a target made up of zircaloy rods and directly cooledby D2O, we plan to develop and demonstrate a target safety container which should be able toprovide tight enclosure of any radioactivity even if something goes wrong in the target itself. Inits present version this container is a double walled aluminium structure with water coolingwhich will surround the target proper. We will initially use it together with the zircaloy rodtarget and hope to proceed to a target with higher neutron yield within a year by replacing thezircaloy rods by lead filled zircaloy tubes. Comparative expected flux figures for differenttarget materials and concepts are given in Table L So far, thermal cycling tests on the lead-filled zircaloy tubes showed no signs of shape change even after more than ten thousandtransitions through the melting point of lead.

Table 1: Calculated flux levels of thermal neutrons for various target concepts for SINQ. Thenumbers are per mA of proton current on target.

Target System

Pb with weaklyabsorbingcontainer

Pb-Bi with steel-container

W-pIates in Al-container

Ta-plates in AI-container

Pb-rods inzircaloy-tubes

Zircaloy

Maximumthermal

(cm"2s-l)

2xl014

0.9 x 1014

0.8 x 1014

0.55 x 1014

1.5 x 1014

0.8 x 1014

Thermal flux at25 cm (cnrV1)

1.3 x 1014

0.85 xlO14

0.6 x 1014

0.45 x 1014

1 x 1014

0.59 x 1014

Loss factor at25 cm radius

1

0.65

0.46

0.35

0.70 (estimated)

0.45

Engineering work on the mark 1 target is virtually complete and construction of the first modelis in progress. Since, according to Table 1, more than a twofold flux gain can be expected froman optimized massive liquid metal target relative to the zircaloy target, development of thisconcept will remain a focus of future work for SINQ. Fig. 4 shows schematically the twotarget concepts.

5. Target services and ancilliary equiqment

The decision to go from a liquid metal targt with no cooling water in the proton beam to asolid target with D2O coolant that will be exposed to the protons had a profound effect on thecooling circuits and other ancilliary equipment. Not only did we have to provide for a delaytank very near the exit of the coolant from the target, but the whole circuitry became

102

i cooling Secondarycoolant

"̂ ^^ I Xi-̂ -X

Figure 4: Schematic representation of thesolid rod target (right) and the liquid metaltarget (left) for SINQ. The solid rod targethas D2O coolant in the proton beam andrequires massive shielding in the forwarddirection of the beam. The liquid metaltarget is self shielding and heat istransported away from the reaction zone byconvective motion of the target material.

significantly more complex and voluminous. With most of the basic engineering for the coolingcircuits now at hand, we are presently establishing a complete 3-dimensional CAD-model ofthe whole system to ensure optimum placement of pipes and equipment throughout. Sufficientspace must be allowed for maintenance and exchange where necessary, so work can beperformed at minimum radiation exposure of personel. The requirements to the targettransport flask have been specified and a conceptual desgin has. been worked out. Detaileddesign is due to start soon.

Similarly, the concept of the control and safety system has been layed down and individualparts of the system are now being detailed.

All pertinent information is stored in a data base on the PSI central computer and is accessibleform everywhere in the laboratory. The data base mangagement system used is ORACLE.

6. Expected neutronic performance

Neutronic performance predictions for SINQ are based to a large extent on model calculations[3], [4], [5]. They were benchmarked against exploratory experimental work done during thepreparatory phase of the project. The expected distribution of unperturbed neutron flux in theD2O moderator is shown in Fig. 5 for the case of the optimum target. For other targets theisoflux contours may not simply scale with the figures given in Table 1 due to a varying degreeof neutron absorption in the target and resulting flux depression close to it. For the thermal andhigher energy groups the flux is given as a function of the radial distance from the target centrein Fig. 6. Fig. 7 shows the expected energy spectrum at 30 cm from the target centre line. Thisspectrum ressembles very much that of a modern reactor, but it should be noted that it extendsall the way up to 500 MeV, although with an increasingly steep drop.

103

The high energy component of the neutron spectrum is not only responsible for the massivetarget shielding but is also a matter of concern in the beam tubes of SINQ. Fig. 8 gives thevariation of its calculated intensity in the beam tubes as a function of the angle of the beamtube relative to the direct line of sight on the target. Obviously the 90° viewing angle chosenfor the beam tubes of SINQ is of great significance. Nevertheless we plan for rather massivemonochromator shielding for the first two instruments in the target hall (see section 8).

7. Experiment infrastructure

In parallel to work on the target block and the target itself, great emphasis is put on thosepieces of equipment, which we call the experiment infrastructure for the instruments. Thisincludes in particular• the cold D2 moderator« the neutron guide system• the beam port inserts• equipment needed to install and maintain these components

Since these systems will serve a large number of instruments over a long time, we consider itparticularly important to provide a state-of-the-art standard, even if this limits the number ofinstruments we can provide for the start-up phase of the source.

Êp5•oc$Eosomg

oc«Inb̂ _ ^~<5o

I

100-1-

80--

60-.

40--

20-

0-

-20-

-40-

-60-_

-80-

""fcooo« » o „ „ %

<o «**fcoQ °^ "•••^. °°> . •*•»0. „ °"„ \»•xr° <v> °« °« °°

•̂i'' « * » *• ^o % q ^

55*\\\\ \ \ \ |

\ } } j ) 1 I 8 I cm2/sec s

*?://•/' / y / 1* ** jt* ° jf° ^ *'-.'* «° 0°° <P° 0°

S* 0* 0^ 0*

rf» 0* 0* *P 0

rf» rf»« ° <pv°

0 20 40 60 80 100Distance from Target Axis [cm]

«S1

1013-

io12-

ion-

io10-

i n « .

\ ^^-^^^

^ \ Thermal

\ \

^ \ \ Intermediate"--v \ \ 1.5eV-*0.82MeV

Fast V0.82— -1 5MeVx \ ^ithermal

\ 0.625— 1.5eV

N

\High Energv>15MeV

20 30 40 50Distance from Target centre-line

Figure 5: Contours of equal unperturbedthermal neutron flux in the SINQ-moderator tank at 1 mA proton currentand for a massive lead target with lowabsorption container.

Figure 6: Radial dependence of theneutron flux for different energy groups inthe moderator tank of SINQ (same targetas in Fig. 5).

104

R=30cm

3500 at R=30cm11500 at R=40cm

10s

10'310-210'1 1 10 102 103 104 10Energy [eV]

Figure 7: Expected energy spectrum in theSINQ moderator at 30 cm from the centreline of the target.

T 10a--

•s- io7-

10°

10 30 50 70Angle 'a' [deg]

Figure 8: Variation of the high energyneutron component seen at the end of abeam tube with the angle of the beam tuberelative to the direct line of sight onto thetarget.

The cold D2 moderator system is presently being procured for an off-line test setup. Thedeuterium-helium heat exchanger is nearing completion, the cryogenic plant is beingcommissioned and the horizontal leg of the cold moderator system including the moderatorvessel and the zircaloy pressure tube are being manufactured. We expect to start the test setupin the second half of this year. The planning of the necessary control system is progressingwell.

Table 2: Specifications of neutron guides at SINQ. Guide coating ist with superrnirrors unlessindicated otherwise (NiC).

Guidedesignation

1RNR111RNR12

1RNR131RNR141RNR151RNR16

1RNR17

Anlge relativeto centre line

of bundle-6°

-4.8°

-4.0°+3.2°+4°+6°

46°

Dimension toWidth x

Height (mm2)50 x 12020x120

30x12035 x 12030 x 12050x50

two times24.5 x 50s

curved lengthLcurv (m)

2020

20202020

24

Radius ofcurvatureRO,™ (m)

14453612

2408206324081445

1234

critical wavelength X* (A)

2.41.7

(NiC)1.51.71.54.2

(NiC)

105

o0\

tTReflectometer

. . l _ i T I M TT......rUZT

Small angle scatteringi___.

Tripleaxis

Test

Polarizedtriple axis

Diffuse elastic

Opticalbench

High res.time of flight

Figure 9: Floor plan of the neutron guide hall showing the arrangement of neutron guides and projected instrument locations.

35000. -

30000. -

25000. -

20000. -

15000. -

2 10000. -

5000. -

Figure 10: Transmissionthrough a 25 cm long and 0.4mm wide "microguide" coatedwith supermirrors designed form = 2 as a function of tilt anglerelative to a well collimatedbeam of 4 A neutrons.

-i.oo -0.50 0.00T i l t angle r

0.50 1.00

One of the six neutron guides viewing the cold D2-moderator will be divided in its height earlyon and the two guides of reduced cross section ( 5 x 5 cm2) will continue with differentcurvatures so that actually 7 guides will penetrate the shielding of the neutron guide cavern(Fig. 9). The properties of the guides are listed in Table 2.

While the two guides which will serve instruments with narrow ingoing collimation will becoated with carbon or nitrogen-doped nickel, we intend to coat the other 5 guides withsupermiirors with twice the critical angle of natural nickel (m = 2). This will be done at PSI byour own staff. Tests have now routinely resulted in supermirrors with m close to 2 andreflectivities of 92% or more. An example of the transmission measured at the reactor S APHDR.with a narrow gap "mircroguide" as a function of tilt angle relative to the beam is shown in Fig.10. Series production on polished glass plates will start in early April. Fig. 11 shows thecalculated neutron wavelength spectrum for the D2-moderator for a 6 m long beam tube of 120x 35 mm^ cross section together with those for the end of a 50 m long guide of the same crosssection coated with natural nickel at 99.5% reflectivity and m = 2 supermirrior at 90%reflectivity. The gain from using the supermirrors is obvious; we expect to exceed the 90%reflectivity on our guides.

1 I ' I ' I ' I ' I ' I ' I ' I ' I 'beam tube 120x35 mm2

5 6 7 8 9 10

Wavelength (A)14 15

Figure 11: Calculatedperturbed wavelength spectrumfor the SINQ cold moderatorfor a beam cross section of 120x 35 mrr?. The dash-dotted lineis the single Maxwellianspectrum for the unperturbedcase shown in réf. [1]. Thelength of the beam tube is 6 m.Also shown are the spectra fora 50 m long guide coated withnatural nickel (m = l, R =995%) and with supermirrors(m = 2,R = 90%).

107

ooo

Beam height delimiter drum-»<£* 'V High engery shutter drums \ Collimator drum

\*V s ^

SINQ-beamhole inserts

Figure 12; Conceptual layout of the beam port inserts with a nine drum high energy shutterthat can also limit the width of the beam. Two user defined drums with horizontal axes can berotated indepently. In contrast to the situation shown, the penetrations in these drums will bearranged in such a way, that they only coincide on the beam axis and are azimuthallydisplaced from one another in the other positions. The first of the two drums is intended tolimit the height of the beam and the second one to hold different collimators.

The beam port inserts in the target block (Fig. 12) will comprise high energy shutters with ninesynchronously driven drums with vertical axes and tapered slits that also allow to limit thelateral width of the beam. The beam height can be limited by another drum with horizontal axisthat has three different open and one closed positions. A second, similar, drum in series withthe first one will serve to hold collimators of different specifications. This provides a very highdegree of flexibility in adjusting the beam on the monochromator to experimental needs and toavoid unneccessary activation of the monochromator environment by limiting the cross sectionof the beam early on. Detailed design work is in progress on these components.

All equipment needed to install and remove the components in the beam ports or coldmoderator insertion penetrations remotely has largely been designed and is almost ready fortender.

8. Neutron scattering instruments

Fully equipped, SINQ could provide space for about 20 neutron scattering instruments even if,as presently envisaged, one of the thermal beam ports will be used to install a sampleirradiation facility.

Apart from the fact that our available resources would not allow to provide instruments for allpossible locations for day one, there are other reasons why we would not want to do this:

• experience has to be gained with the situation on a cw-spallation source, for which there isno real precedent world-wide;

• we feel that it is important to have a basic set of well designed and equipped standardinstruments which are fully understood and can be used for science from the beginning,continuing the tradition that exists at our institute;

• we wish to reserve space for advanced developments which currently exist only asconcepts or proposals, but which take more effort to implement than we can presentlyprovide while the whole source is under construction.

According to current planning, there will be a minimum of five instruments ready when thesource will be commissioned, with some hope that three more could become operationalshortly afterwards with external support In the following we give a brief description of the fiveday-one instruments, which are summarized in Table 3.

A high resolution powder diffractometer (HRPD) similar in design to the DMC presently in useat SAPHIR will be located at one of the thermal beam ports. The position sensitive detectorcovering an angular range of 160° can be rotated around the sample. A 20 cm high verticallyfocussing Ge-monochromator will illuminate a cross section of 2 x 5 cm2 at the sampleposition. There will be two fixed monochromator angles at 2 0M = 90° and 120°.

HRPD will become a versatile, powerful tool to investigate both structure and magneticordering phenomena in a large class of materials. The applications range from solid statephysics, chemistry, crystallography, materials science to biology and include both fundamentaland applied research of interest also to industry (e.g. to refine crystal structures of lowsymmetry for unit cell volumes up to about 2000 A3, as a function of temperature in the rangeof 7 mK to about 1400 K with up to approximately 250 parmeters). Also zero matrix pressurecells, superconducting magnets etc. should become available for similar powder investigations

109

Table 3: First generation instruments planned for SINQ under PSI-responsibility

Instrument

location

monochromator

•"- "mon

detectors

resolution/range

spécialfeatures/remarks

High resolution powderdiffractometer

thermal neutron beamport

Ge(hkl)

20 + 90deg

position sensitivedetector 10 - 170°

Ad/d = 4-10'4

vertically focussingmonochromator;GdO mylar sollers infront of detectors

Four circlediffractometer

thermal neutron beamport

PG (002), Cu(220)

15 - 90 deg

2 dim, He-3210x1 85 mm2

(microstrip)

dilution cryostat;closed cycle cryostat

High resolution tripleaxis

cold neutron guide(s.m.)

PG (022), Be (220)

30-145 deg

He-3

>7ueV

Polarized neutronstriple axis

cold neutron guide(s.m.)

Heusler(lll)PG(002)

30-145 deg

He-3

>7ueV

Small angle scatteringinstruments

cold neutron guide

mech. velocity selector

„

2 dim,128 x 128 cells960 x 960 mm2

S-lO^nm-^Qi.lOnm~l

double crystalmonochromator(option)polarization (option)specimen changercryostat;furnace;electromagnet

High resolutingTime-of-flightspectrometer

cold neutron guide

chopper 3-12 A

„

He-3 detector array

0.005 f 0.5 meV

counter rotatingdisc choppers100 1- 400 Hz;detector angularrange 10 -135°;converging supermirror

in the future. Depending on the particular problem, the amount and quality of the sampleavailable, resolution and intensity may be optimised.

The second beam on the same beam port will be occupied by a four circle diffractometerequipped with three 2-dimensional position sensitive detectors which are mounted onindividual columns and can be moved around the sample on a common support plate. Since weintend to provide sample environment equipment for a rather wide range of conditions,including an ADP closed cycle helium refrigerator for 20 - 450 K and a dilution cryostat for~200 mK, tilting options for the sample may sometimes be limited. For this reason it will bepossible to move the detectors out of the horizontal plane individually. 20 x 25 cm2 microstripdetectors with a resolution of 1 to 1.5 mm horizontally and 3 mm vertically will allow to obtaindetailed information on the intensity distribution of individual reflections if desired. Variationof the sample-detector distance will be possible by manual adjustment. In addition to singlecrystal diffraction with the option of investigating the 3-dimensional intensity distribution ofindividual reflections by the 2-dimensional detector plus rotation of the sample, the instrumentcan of course also be used for texture studies with great benefits.

Fig. 13 shows the layout of the monochromator shielding and Fig. 14 gives a schematicrepresentation of the four circle diffractometer.

variable angle monochromator shieldfor 4-circle diffractometer

Figure 13: Design of themonochromator shieldingat the thermal beam panoccupied by a highresolution powderdiffratometer and a fourcircle diffractometer. Toshield also against highenergy neutrons, asandwich-type shielding ofheavy and moderatingmaterial was chosen.

2 fixed anglesfor HRPD-monochromator

Two triple axis spectrometers of similar design are planned for installation on the neutronguides. One will be a conventional high resolution instrument mainly for small energy transfersand the other one will be equipped with a Heusler crystal for polarized neutrons.

A schematic representation of the design of the triple axis spectrometer is shown in Fig. 15Great care is being taken to keep the background in the detectors as low as possible byopening up the shielding of the monochromator and analyser as the tables move only in suchplaces where it is faced by other shielding parts.

Ill

Figure 14: Schematicelevation of the four circlediffractometer showing thedetector in two differentlyinclined positions.

The non-polarized triple axis spectrometer will be used mainly for the investigation ofstructural phase transitions and high resolution spectroscopy. The super mirror coating of theguide will alow to use incoming energies between approximately 2 to 3 meV and to reach amomentum transfer Q up to about 5 A'1.

Detector

Sampleta

Figure 15: Conceptual layout of the triple axis spectrometers at the SINQ neutron guides.

112

While principally of similar design, spin flippers and guide fields along the neutron flight pathswill make the TAS for polarized neutrons somewhat more complex. Apart from the obviouspossibilities to use also this machine in a non-polarized fashion, curved Heusler crystals orpolarizing super mirrors will be available as monochromators or analyzers, depending on theneutron energies used.

Standard components will be used to build the sample table for all four of the aboveinstruments. They will be movable on air cushions and have compatible supports forstandardized sample envrionment equipment. The clearance from the sample table to the beamcentre Une has been set to 400 mm. The spectrometer arms are designed for 0.01° positioningprecision. The precision of angular positions of the crystals has been specified as 0.005°. Forall four spectrometer the standard components such as air cushion tables and turn tables havebeen decided upon and are being purchased. The position sensitive detector for the HRPD hasbeen ordered and individual parts for all spectrometers are being designed.

The fifth instrument planned for early operation is a small angle neutron scattering instrumentwith a maximum sample to detector distance of 20 m and a neutron guide-collimator

Sample position tablewith x-y-<]> adjustment

Evacuated flighttube for scatteredneutrons with built-indetector displacement gear

Mechanical velocity selectorwith adjustable tilt angle andlateral position

Laterally movable 2 dimensional positionsensitive detector with independentlyadjustable beam stopper

Collimator-neutron guideexchanger with individuallymovable sections of different len&ths

Figure 16: Main components of the SINQ-SANS instrument.

interchanger that allows to match the collimation in front of the sample to the resolution at thedetector. It will be equipped with a mechanical velocity selector but could alternatively also usea double crystal monochromator by taking advantage of the fact that two vertically displacedprimary beam paths are possible in the neutron guide-collimator exchanger. This might be ofpraticular interest, if polarized neutrons are desired. The 124 x 124 element 2-dimensionalposition sensitive detector can also be moved sideways inside the large vacuum vessel toincrease the accessible Q-range. The beam stop in front of the detector can be adjustedindependently.

113

The main components of the SANS-instrument are shown in Fig. 16. With the exception of theelectronics for the data acquisition system and instrument control, all of the main componentshave either been ordered or are out for tender at present.

In addition to these five instruments which are designed and looked after by PS I staff, we arepreparing to accept at least three more spectrometers for which interest has been expressed byexternal groups. These include a neutron-optical bench with several sets of pairs of perfectcrystals for neutron interferometry, very high resolution small angle scattering and relatedwork, a back scattering spectrometer and a high resolution time-of-flight spectrometer. If theseprojects ma ture, SINQ can make available a full suite of high perfomance instruments to itscustomers and users, while more space is available for new developments.

9. Concluding Remarks

With the construction and operation of SINQ the Paul Scherrer Institut intends to furtherimprove and enlarge its comprehensive set of tools for condensed matter research which alsoincludes muon and pion beams and possibly even an intense positron beam facility in the nearfuture, if an ongoing development turnes out successful. This is in accordance with thelaboratory's shifting emphasis towards condensed matter and materials research. More longterm plans aim at providing also a synchrotron light source.

During the planning, design and construction of our facilities we have received help and advicefrom many places and in many forms. We will be glad to pass our own konwledge andexperience on to others as the need arises. PSI's facilities are open to outside users undervarious schemes of collaboration which can be agreed upon individually.

References

[1] O.S. Bauer, SINQ-status report, ICANS-XI, KEK-Report 90-25 (1991)

[2] G. Heidenreich and O.S. Bauen "The PSI-SINQ Target Development Programme" in PSI-proceedings 92-03 (1992), p. Cl

[3] F. Atchison: "Neutronic Considerations in the Design of SINQ": PSI-proceedings 92-03(1992), p. Bl

[4] F. Atchison et al.: "The D2 cold neutron source for SINQ" in Advanced Neutron Sources1988, InsL of Physics Conf.Ser. Nr. 97 (1989)

[5] F. Atchison: "Data and Methods for the Design of Accelerator Based TransmutationSystems", PSI-proceedings 92-02 (1992), p.440

114

NEUTRON SCATTERING RESEARCH WITHLOW TO MEDIUM FLUX REACTORS

J.J. RHYNEResearch Reactor Center,University of Missouri-Columbia,Columbia, Missouri,United States of America

Abstract

This paper discusses neutron scattering instrumentation that is appropriate for research at low to mediumflux reactors. A summary is given of the feasibility and potential for various categories of scattering instrumentsat reactors of different thermal flux levels. Included are suggestions for powder diffraction, small anglescattering, and triple axis spectrometers. A brief discussion is included on methods improving signal to noiseratios, particularly in light water moderated facilities.

1. INTRODUCTION

This paper presents suggestions and guidelines for types of instruments andassociated research programs that are appropriate for reactors in the low tomedium flux categories (i.e., 1 x 1013 < flux < 5 x 1014 n/sec/cm2). The researchand instrumentation examples given are admittedly highly subjective andcertainly not meant to be all inclusive.

Much of the following discussion is based on experience gained by the staff at theUniversity of Missouri Research Reactor Center (MURR) and from the author'sown experience at this facility and at the NBS Reactor at. the National Institute ofStandards and Technology in Gaithersburg, Maryland.

The University of Missouri Research Reactor is a 10 MW research reactor that isthe highest flux reactor at a university in the United States. It is a light watermoderated and cooled, beryllium reflected, pool-type reactor with an annulus flux-trap pressure vessel. Eight MTR-type fuel elements form a 28 cm diameter x 61cm high core and produce a flux of 1.2 x 1014 n/sec/cm2 at the end of the six radialbeam tubes used for scattering experiments. The peak flux in the flux trap is 6 x1014 n/sec/cm2. Three of the beam tubes are 15 cm in diameter and three are 10cm diameter. Extensive use is made of liquid-nitrogen cooled and also roomtemperature silicon and sapphire filtering in the beams before the instrumentmonochromators. This is an essential requirement with light water reactors ofgreater than about 1 MW power level with radial beam tubes, in order to reducethe fast neutron background and to permit monochromator shielding of compactdesign.

The present neutron scattering instrument complement at MURR consists of atriple axis spectrometer, two double axis single crystal units, two powderdiffractometers, and two neutron interferometers. A 15 m small angle scatteringspectrometer is under development as are a new triple axis spectrometer, areflectometer, and a dedicated residual stress facility.

A summary of MURR and other instruments used for powder diffraction, SANS,triple axis spectrometry, and neutron reflectometry will be given in the followingsections.

115

Suggested Capabilitiesof Reactors for

Neutron ScatteringResearch

Operates>100 hours

er week

Education &Training Only

Powder Diffraction

Some Research

Powder DiffractionSANS

Significant Research

DiffractionSANSReflectometry

Major Research

Diffraction, SANSReflectometryInelastic Scatter.

Figure 1. Diagram for suggested types of research and associated instrumentation for reactors with the fluxes given at the top. Theflux numbers are considered more critical than the power figures presented. The flux/power relationship is scaled based on severalH2O and D2O reactors in the U.S. and is characteristic of a compact core design.

2. GENERAL CONSIDERATIONS

In order to have truly competitive and useful research instrumentation at low andmedium flux facilities, great care must be taken in the selection and designofscattering instruments. The first advice is to carefully analyze the real andpractical capabilities of the reactor source and use this analysis in the selection ofappropriate research areas and the required instrument complement. Thenatural tendency is to build too sophisticated instruments or to try too complextypes of experiments that are both beyond the capabilities of the reactor source.This approach is often wasteful and often doomed to failure.

Figure 1 gives a suggested (and highly subjective!!) guide to the selection ofresearch and instrumentation, based on the available source flux for the beamtubes. The table can be summarized briefly as follows: (1) For reactors with fluxesless than 1 x 1013 n/sec/cm2, particularly if they are light water moderated, theuses should likely be limited primarily to education and training rather thancompetitive research. Limited powder diffraction is certainly feasible with thisflux level. (2) Reactors with fluxes above level (1) but less than about 5 x 10isn/sec/cm2 can produce research with powder diffraction and SANS instruments,and, if they operate more than 100 hours/week, could be considered to havesignificant research capabilities. A neutron reflectometer is also a suggestedaddition to the stable of instruments. (3) For fluxes higher than 1 x 10*4n/sec/cm2, significant first-rate research is definitely feasible, and, if the sourceoperates more than 100 hours/week, one can consider such a reactor a majorresearch facility capable of performing diffraction plus inelastic scatteringresearch. It is particularly important if beam tubes, of. greater than 10 cmdiameter are available to use beam focusing monochromator techniques,especially on inelastic scattering spectrometers. [It should be noted that, whilethe examples given in Figure 1 can be considered representative, there are manyhighly successful reactor research programs operating well outside the envelopeof the above guidelines. Clever instrument optimization and good shieldingconcepts, plus patience with long counting times, can make up a lot for lower thanoptimal reactor fluxes.]

It is often the case that lower flux reactor facilities must also contend with highlyconstrained budgets for instrument design and construction. This is mostunfortunate and probably means that building a small number of really first-classmachines is better than doing a mediocre job on a larger group of instruments.In general the advice could be that "if available funds are insufficient to build astate-of-the-art spectrometer ~ wait and argue strongly for the additional money."

3. BEAM AND INSTRUMENT OPTIMIZATION

The necessity to optimize all instrument and beam parameters at every step to getthe most out of the available reactor source flux is a critical guideline forinstrument design at lower flux facilities. The important end result is the highestpossible signal to noise ratio. Minimizing backgrounds often carries the highestpayoff in this regard. The use of fast neutron filtering, appropriate high-densityshielding materials, and adequate care to fully shield all neighboringinstruments in the reactor hall is essential. Pressing of monochromator crystalsto open up the mosaic to the range of 20 minutes and the use of focusingmonochromators to take full advantage of the vertical divergence of the beams canincrease intensity appreciably. As discussed below, for the case of powderdiffraction, multiple discrete detector systems or position sensitive detectors canalso raise data rates by significant factors.

117

Table 1*. FILTER MATERIALS TO REDUCE FAST NEUTRONCOMPONENTS in THERMAL BEAMS

Note: O = Ofast/(öabsorption + Oincoherent) ÏSfast neutrons.

Advantages

Si Good crystals available

SiO2 Good crystals, modest (E

Bi Excellent a

Al2Os High ©D

MgO High ©D

MgF2 Good a

Bi-silicates Excellent a, high dens.

Bii2Si020

Alexandrite Good crystal qualityBeAl2C>4 Modest a

Cubic Good crystals, highZirconia high density

• Compiled byW. Yelon (MURR)

relative cross section for

Problems

Needs cooling, poor 7reject.Average a

Poor crystal quality,low ©D •Oxygen window -2.35 MeVLarge crystalsunavailable,

oxygen windowPoor crystal quality,low ©DPoor crystal quality,low ©n

Small crystals only

O-window, absorption(need Ca

stabilization)

Incident beam filtering with liquid nitrogen cooled silicon, sapphire (may beoperated at room temperature), or alternate materials, preferably before the beamexits the biological shield, is very important as is a shielding design thatminimizes self-generation of gamma rays. If only longer wavelengths are needed(e.g., > 3.96 A) for SANS, a combination of Be and Bi is a very effective filter. Table1 gives a list of candidate materials for fast neutron filtering prepared by W.Yelon1. The general criteria for a good filter are a material with (1) a relativecross section for fast neutrons [o = Ofast/(aabsorption + «^incoherent)] of order 20-50, (2)good crystal perfection to reduce losses from Bragg scattering, (3) a high Ooebye toreduce phonon scattering, (4) reasonably priced and available single crystals, and(5) a high density to provide good y-ray rejection. As an alternative or in additionto filtering, beam offsets using a double monochromator design are also veryeffective at reducing background, but may extract a significant penalty also inprimary beam intensity depending on the monochromator crystals.

118

Finally neutron guides are an optimal approach to better signal to noise ratios asthey allow instruments to be located well away from the reactor face. Ifsupermirror coatings are employed, guides will function extremely well even inthe traditional thermal (not cold) neutron wavelength range.

4. DESIGN OF SPECIFIC NEUTRON SCATTERING INSTRUMENTS

The following are some suggested design considerations for neutron instrumentstargeted for use on lower flux reactors. These guidelines are very cursory innature and details must be obtained from review articles or via communicationwith laboratories having completed similar instruments.

4.1 Powder Diffraction

The powder diffractometer frequently becomes the workhorse instrument at manyreactors and often is the instrument (along with SANS) in greatest demand due toits applicability to a variety of problems in fields ranging from materials science,to biology, to complex chemistry. Powder diffraction research also necessitates theuse of a modern Rietveld analysis computer code, generally capable of handlingboth multiple material phases and magnetic scattering, such as GSAS(Generalized Structure Analysis System) written by the Los Alamos NationalLaboratory scattering group.

For powder instruments, to optimize resolution, the monochromator take-offangle should be designed to be large (preferably & 90°) so that the resolutionminimum falls at large scattering angles where the density of reflections ishighest. A wavelength of approximately 1.5Â is often selected, and manyinstruments incorporate variable wavelength or multiple monochromators. Forcutting edge research, Soller slit collimations should be of 101 or less throughout(before and after the monochromator and before the detector).

He detec tors

Cryostat

Söller coilinrvctors(5'-30')

Shielding

Figure 2. Schematic design for a powder diffractometer using multiple discrete detector tubes.

119

There are a variety of detector designs in use at various facilities. A singledetector tube design is most basic, but today can really only be considered adequatefor educational purposes. Multiple detector tube designs incorporating 5, 32, 64 ormore discrete detectors, each with its own Soller collimator, allow simultaneouscollection of data from multiple portions of reciprocal space. A diagram of such a

A schematic representation of thedetector shield assembly

40.5cm

Rgure 3. Schematic design for a powder diffractometer using a stacked array of linear positionsensitive detectors. This machine is in operation at the University of Missouri ResearchReactor Center.

multi-detector is shown in Figure 1. Rotating the entire assembly undercomputer control through an angle slightly greater than the detector angularseparation provides a seamless data set for analysis. For arrays with a smallernumber of detectors (e.g., 5 tubes placed 20° apart), larger angular movementsmay be necessary to adequately cover the full scattering range. The multipledetector instrument is a very powerful research tool if combined with a properlychosen monochromator, a high 26mono and very good angular Soller slitcollimation (101 or tighter). A complication is in the calibration of the multipledetectors, which generally all have discrete zero corrections and detector countingefficiencies. Calibrations must be periodically checked to ensure thatdiscriminator or electronic gain drifts are not a problem that can affectbackgrounds.

Position-sensitive detector arrays have provided a major advance in powderdiffraction especially in the speed of data acquisition. The most sophisticated isthe "banana" detector in which the continuous detector element is curved atconstant radius about the sample. This detection scheme can only be consideredby facilities with very opulent instrumentation budgets. A much more modestapproach (shown in Figure 2) is the use of commercial linear position-sensitivedetector tubes, either singly or in a multiple vertically stacked array, to providecontinuous data accumulation over a sector of the circle .about the sample.

120

Multiple segments may be used or the array rotated to cover the full range ofreciprocal space. The correction for the detector being a sector instead of a truearc can be made in the data acquisition software. Such an instrument,incorporating five stacked linear detectors and using electronics designed by R.Berliner of MURR, provides excellent resolution and stability, and has been incontinuous use for several years. Increasing the number and height of thedetector stack can further increase data collection rates, but a Debye Sherrer conecorrection must be applied via software. A minor disadvantage to this type ofdetector is that resolution is dependent on sample size (and detector to sampledistance) and an oscillating vane collimator is needed to suppress spuriousincoherent backgrounds from cryostats, furnace shields, etc.

The conventional monochromator and Soller slit configuration can be replacedwith a perfect crystal two-dimensionally focusing monochromator as developed byPopovici et al.2 This remarkable and simple breakthrough uses ahemispherically bent perfect crystal (e.g., Si) to provide a significant gain inresolution over conventional mosaic monochromator s. Peak width is typicallyreduced by a factor approaching three at high scattering angles with asimultaneous gain of three in peak intensity. The gains can be made largest atthe high scattering angles where good angular resolution is most critical.

4.2 Small Angle Scattering Spectrometers

Sharing the limelight with powder diffraction, small angle neutron scattering(SANS) is a technique in great demand for research in a broad cross section ofdisciplines from materials science, to polymer chemistry, to biology, andmagnetism. SANS bridges the real space dimensional gap between wide anglediffraction (interatomic spacings « 5-10Â) and conventional microscopies in themicron size range. Modern SANS spectrometers typically cover a range or wavevector transfer 0.0005 Â-* < Q < 0.5 A-i and find applications for examining defectproperties, large molecule structures, local magnetic distributions, phasetransitions, and critical and subcritical scattering in disordered systems amongothers. As shown in Figure 4, a conventional SANS machine includes a pre-filterof Bi-Be to screen out neutrons with wavelengths less than 3.96 A and unwantedY'S. A wavelength appropriate for the resolution and other parameters desired isselected by a broad band (AXA- « 10-20%) variable wavelength monochromatingdevice such as the helical velocity selector shown. Neutrons diffracted at smallangles from the sample are detected on an area detector (continuous wire,discrete wire, or micro-strip) providing a two-dimensional representation of thescattered neutrons.

SANS instruments are among the more expensive machines to build, largely dueto the size of the shielded evacuated flight paths and the cost of the area detectorand velocity selector (or crystal monochromating devices.) Horizontal space fromthe reactor face to the outside wall of the confinement building is also a problem inmany facilities, unless a penetration is available. Useful spectrometers must be atleast 8-10 m in length.

An alternative is to construct a vertical flight path in which the beam from thereactor is deflected upward 90° by a crystal monochromator (e.g., pyrolyticgraphite) and then the detector and flight path are located on a tower. The verticaldesign has some severe drawbacks, however, among them (1) the difficulty ofachieving a sufficiently large AX/X from the monochromator, (2) lack of variablewavelength, and (3) the difficulties of accommodating cryostats, magnets, and

121

M

tlBismuth- Velocity First Collimating

Rotating Shutter Beryllium Selector ApertureFilter _ . , - , . ,Evacuated Flight Path

SampleChamber

Evacuated Flight Path

Area Detector

Figure 4. Drawing of a typical SANS machine using an helical velocity selector as beam monochromator and a 64x64 cm areadetector.

other sample environmental equipment. A vertical SANS facility does offer greatadvantages in the study of liquids, however. For both horizontal and verticalconfigurations, it is necessary to provide an evacuated chamber for the areadetector that allows a variable sample to detector distance to control (along withthe variable wavelength) the q resolution (minimum q) and to vary the total rangeof q accessible on the detector face.

4.3 Triple Axis Inelastic Scattering Spectrometers

Triple axis spectrometers are among the most demanding of precise attention todetail and careful design of any of the instruments discussed here. As statedabove, they also require fluxes in or above the 1 x 1014 range to be scientificallycompetitive. Elaborate care must be exercised in the design of the shielding toreduce the intrinsic background and also the thermal neutron and y backgroundgenerated in the shielding itself from fast neutrons. Well-designed instrumentscan achieve background levels (beam open without sample) of a fraction of 1 countper minute. Building both monochromator and analyzer assemblies for variableenergy transfer also significantly increases the cost of the instrument, but thisfeature is highly desirable for appropriate flexibility in selecting resolutionconditions. In addition to filtering the white beam against fast neutroncontamination, filters (usually again of costly pyrolytic graphite) are needed in thefixed energy (monochromator or analyzer) scattered beam to remove X/2 andhigher order contaminations. In general, the decision to build a triple axisspectrometer at a low to medium flux reactor should be made only after carefulthought and planning and examination of such instruments at comparablereactor sources.

4.4 Neutron Reflectometers

Neutron reflectometers are among the newest instruments at most reactorfacilities and have recently enjoyed wide success for problems in magnetism,polymeric films, and biology. Essentially, the reflectometer is a two-axisdiffractometer in which the monochromatic beam is narrowed by slits to typicallymuch less than 1 mm. The outgoing beam, scattered at grazing incidence, is thendetected by a well-shielded conventional detector. Because of the small scatteringangles involved, elaborate steps must be taken to minimize backgrounds and toremove higher order wavelength contamination. In a typical experiment, thedata at the lowest Q represent neutrons that are totally externally reflected whichdefines unity reflectivity. Above the critical angle the reflected neutron curve fallssharply in intensity. It is in the analysis of the Q-dependence of the reflected dataabove ©crit that the physical information on the details of the interface (magneticmoment or density gradient, etc.) are obtained. The field of neutron reflectometryis currently attracting a lot of attention and is an instrument that should beseriously considered by a reactor facility if sufficient flux is available and lowbackgrounds can be achieved.

5. SUMMARY

In this paper we have tried to present some rudimentary details andconsiderations for developing neutron scattering instruments at low and mediumflux reactors. Many of the conclusions and recommendations are admittedlysubjective and far from comprehensive. Because of their ubiquitous nature, theonly instruments considered for this discussion were powder diffraction, SANS,

123

triple axis, and refîectometer machines. There are other types of scatteringinstruments, generally more specialized, which have been built and verysuccessfully used at many intermediate flux sources throughout the world.Polarized beam scattering instruments were also not considered here becausethey are more demanding of source flux (with the exception of depolarizationexperiments).

As a first line of information for reactor facilities contemplating new orupgrading scattering instruments, the many articles in Neutron NewsS can bevery helpful and point the way to more extensive references concerning details ofspecific instruments.

REFERENCES

1. prepared by W. B. Yelon (unpublished.)

2. M. Popovici and W. B. Yelon, "Neutron Focusing with Bent Crystals," Zeitschriftfür Kristallographie (to be published in 1993); M. Popovici and W. B. Yelon,"Design of Microfocusing Bent-Crystal Double Monochromators," Nucl. Inst. andMethods (to be published in 1993).

3. Neutron News, published quarterly by Gordon and Breach SciencePublishers, Reading, Berkshire, UK, and the companion Journal of NeutronResearch.

124

THE ORPHEE VERSATILE LOW-COST MULTIPROCESSORSYSTEM FOR DATA ACQUISITION AND CONTROL OFNEUTRON SPECTROMETERS

G. KOSKASLaboratoire Léon Brillouin,CEN Saclay,Gif-sur-Yvette, France

Abstract

This paper describes the new data acquistion and control system ofthe neutron scattering instruments at the ORPHEE research reactor.The existing system has undergone a complete change: the original CAMACsystem and minicomputer controlling each experiment have given way tocommercial CPU boards and microcomputers like the IBM PC. The com-munication links between these two components are the IEEE 488 or RS232standards. Emphasis is placed on the flexibility and modular nature ofsuch a system which makes a maximum use of commercial products thusguaranteeing reliability and ease of use. A study of the requirementsand evolutions, technical as well as philosophical, is detailed todemonstrate the motivation of the choice of the system architecture.A survey of the various hardware and software achievements and finallyan overview of future improvements is given.

INTRODUCTION

There are 23 neutron spectrometers installed around the ORPHEEresearch reactor at the CEN-Saclay (France). As the instruments wereconstructed, their hardware and software became extremely varied anduncoordinated. Most of the experiments used the CAMAC system and amultitasking minicomputer shared by several users, while others usedspecialized calculators (DIDAC, MULTI 20) or a centralized micropro-cessor system. The CAMAC included a number of different modulesdedicated to each kind of command or measurement. The physicists werenot able to write, on the driving computer, the software required byeach module, so their handling was reserved for programmers only. Inaddition, k or 5 layers of serial electronics were introduced betweenthe computer controlling the experiment and the physical measurement orcontrol. This complexity made tests and repairs extremely difficult.With such a diversity in material and design, problems arose concerningmaintenance and implementation of new instrumentations. A newacquisition and control system was therefore conceived, bearing inmind that most of the existing spectrometers had to be reused and thatthe new system should be easy to maintain, expand and control. Theserequirements are similar to any neutron and X-ray spectrometer ofmedium rate data acquisition. The system architecture described belowis a solution in updating any reactor or synchroton instrumentation andin meeting new requirements.

ELECTRONIC REQUIREMENTS OF ORPHEE SPECTROMETERS

The requirements are similar to those of process control. Unlikeneutron spectrometers for pulsed sources, the time of flightinstrumentation is a small part of the entire facility ( 1/4 at ORPHEE)

125

and acquisition rates and time resolutions are 20 times less. Ourspectrometers (like triple-axis or two axis) are machines where controlfeatures are as essential as acquisition features because theexperiments are not static: the experimental environment must changeautomatically (wave length, temperature, pressure, magnetic field,sample etc...).

Even though our 23 spectrometers vary considerably they share acertain number of functions:PULSES COUNTING AND TIME COUNTING

The duration of an experiment varies from a few minutes to severaldays. Two to ten counters are used with preset sealer or timeroperation. The time resolution is 1/100 sec. wide and the maximumcontent of the sealers is more than 10,000,000 counts. There are 2modes of operation: continuous mode and "flipping" mode where neutronsare alternatively submitted to a polarizing field.POSITION SENSITIVE DETECTOR (P.S.D.) ACQUISITION AND TIME OF FLIGHT

A P.S.D. can be assimilated with a mosaic made of counters or"cells" (30 to 16000 at ORPHEE). Discriminator modules and positiondigitizer modules are already integrated in the P.S.D. electronicswhose role is to send the number of the cell hit by the neutrons. Thesimple non time of flight mode consists in counting the number of timeseach cell is touched. The time of flight operation is considerablysimplified by the performances required in acquisition rate (highestrate: 50 KHZ ), time resolution (a few usée) and maximum time channelnumber (250). Nevertheless, in order to reduce the amount of data, itis necessary to group cells into sets of various sizes and shapes.MOTOR MONITORING

Some spectrometers have more than 15 motors (D.C. or stepping) .Their power ranges from 1 to 100 Watt typically. The positionningprecision (1/100°) requires a closed loop feedback control.SIMPLE TTL I/O

TTL I/O are used to monitor functions like position readings, D.C.currents, temperature control, relay switching etc... These equipmentsexist in commercial systems and are conceived generally to be monitoredby about 50 TTL I/O bit (e.g. position encoding, D.C. supplies).ABANDON OF CAMAC SYSTEM

CAMAC was not suited to our instruments because CAMAC modules wereeither overdimensionned, inadequate or inexisting.

Pulse counting in the "flipping" mode needed 5 commercial and 1customized CAMAC modules. Commercial CAMAC modules are excellent forhigh data rates in time-of-flight acquisition but overdimensionned forour needs. Such an application generally requires 3 or 4 kinds ofCAMAC modules: histogramming module, TDC (time to digital converter)module, memory module (for data compression), FIFO module. There areno commercial CAMAC units to solve the problems of closed loop feedback

126

control of D.C motors with different kinds of position encoders(resolver, absolute or incremental optical, potentiometer). They haveto be designed, are therefore expensive, and complicated to use becauselogical commands are separated from power commands. In conclusion,CAMAC was not advantageous in the control of our spectrometers because:- CAMAC standard is not commonly used in industrial process controlwhereas our systems can profit from industrial applications (robo-tics, digitally controlled machine-tools, regulation, automatizationetc...)

- Communication with CAMAC units is bulky because it is encoded- CAMAC is expensive because of its poor distribution in industrialfirms

- Computer choice is limited since few computer makers offer interfaceand CAMAC firmware on their products.

NEW SYSTEM DESCRIPTIONThe aim of the system is to create a structure able to reuse the

existing equipment more efficiently, accept commercial devicesimmediately and permit transformation into or conception as"intelligent spectrometry peripherals" of systems.

The basic concepts of the new system are:- Division of the spectrometer into intelligent parallel completelyindépendant electronic functions or "peripherals"

- Undertaking by the peripheral instead of by the computer of allthe necessary firmware for driving each function

- Abandon of the interconnecting bus between electronic boards- Simplicity of the interactive dialogue between the peripheral and thecomputer or the manipulator

- Unicity o'f the CPU board used in the stand alone mode to drive orfulfill each function

- Realization of the function by software rather than hardware- Reduction of the number of interfaces between the functions andthe computer

- Driving of the spectrometer by any kind of computer- General use of widely marketed products- Adoption of the IEEE 488 and/or RS232 standard to connect eachperipheral (or CPU board) to the computer

To understand the system structure, the recently installed SpinEcho spectrometer will serve as an example (fig N° 1). It is composedof two levels:I/ the commanding level and its 2 micro-computers - one commands theinstrument while the other does a first order data processing in orderto optimize the experiment2/ the level that groups the different functions or modules of theinstrument, each equipped with the same CPU board but with differentcustomized firmware. The particular nature of the instrument requiresthe following 6 modules :

127

ADRESSABLE ZONE (MEMORY AND I/O CHIPS!

MICROPROCESSOR6809 MOTOROLA

CLOCK/CALENDAR58174

\RAM/ROM/EEROM

11 VIA

16522FOR

INTER-NALUSE

VIA6522

VIA«522

VIA6522

VIA6522

TMS 99H ANLFOR IEEE 488 BUS

ACIA 6850FOR RS232

ACIA 6850FOR RS232

EXTERNAL CONTROLLERANY MICRO/MINI/

MAIN FRAMECOMPUTER OR RS232

TERMINAL

EXTERNAL HARDWAREENVIRONMENT

e.g. ANY ELECTRONICINTERFACE

Figure 1

a) the module monitoring the stepping motors (one CPU card can control12 or more motors)

b)angular position measurement modulec)the module monitoring the D.C. regulated power suppliesd)time monitoring and pulse counting module (5 counters)e)fast time of flight with one counter modulefjdata acquisition module with XY 32X32 P.S.D. and time of flight

Besides these peripherals, the spectrometer is equipped with otherIEEE 488 commercial instruments (high performance voltmeter, high powerD.C. supplies ...)

Each module is an IEEE 488 peripheral of the controlling computerand can be directed separately by just one RS232 line for the wholespectrometer. It goes without saying that, whenever a new function hasto be added to the spectrometer, it is enough to connect a new IEEE 488module on the BUS.

NEW SYSTEM ADVANTAGES

The structure is simple and regular and places the commercial andcustomized peripherals on the same level. Being autonomous, the peri-pherals off-load much of the monitoring computer's tasks and severalinstructions can be given simultaneously (eg counting and moving) dueto the parallel structure. In addition, separating the spectrometerinto intelligent and indépendant functions follows the physical featu-res of the instrument. The users (physicists, biologist, chemist) caneasily understand the system design and technicians can repair it

128

quickly. They can, for example, communicate with each part of thespectrometer by using a simple mnemonic language without passingthrough the controlling computer or disconnect momentarily a function(e.g. sample changer) without disrupting the experiment. The modularnature thus reduces further developments and brings uniformity, forthough each spectrometer is specific, it is composed of a subset ofcommon functions.

Due to the integration of components and electronic functions, allthe features needed to solve each problem (CPU, memory, TTL I/O ports,serial I/O ports, A.D. and D.A. conversion, communications etc...) cannow be set on a single board. From now on, the interconnecting bus canbe abandonned and immediate advantages are obtained:- no central controlling of the bus (which blocks the whole system incase of break down)

- Total independence of the CPU boards; their programming thereforebecomes simple

- simplicity in setting up due to the independence of each function- gain of speed and reliability due the decrease in interfacing devices- no physical restraints of connecting the board on the same crate orlinking différents crates

- possibility of integrating the CPU board into existing devices toconvert them into intelligent computer peripherals, just like thosein industry.

The reasons for choosing software rather than hardware weredetermined by several observations :- The microprocessors available on the commercial market enable, inmany cases, a microcoded electronics to compete with a previouslyhardwired one in speed of execution, and does more.

- Software design outclass hardware design in flexibility andease-of-operation

- The software conception of a function is cheaper because it entailsonly the conception cost, not the manufacturing.

The IEEE 488 bus (GPIB) corresponds exactly to our requirementsbecause it is the "IEEE STANDARD DIGITAL FOR PROGRAMMABLEINSTRUMENTATION" and is widely used in industry. This communicationbus also provides the "opening" of the structure: for example, ifperformances so required, the system could integrate a CAMAC or VMEcrate equipped with a GPIB/CAMAC or VME interface. Vice versa, ourperipherals can be connected to CAMAC or VME crates equipped with CAMACor VME/GPIB interface.

COMPARISON WITH THE FORMER CAMAC SYSTEM

INITIAL COSTThe setting up of the "SPIN ECHO" spectrometer with the CAMAC

system would have necessitated 11 modules of 8 different kinds:one stepping motor controller ($1500) for function a)one TTL I/O port ($1000) for function b)one TTL I/O port for function c)

129

one clock unit ($1000) andone scale unit $1500) for function d)one histogramming unit ($2000) andone TDC unit ($3000) for function e)one histogramming unit,one TDC unit andone memory unit ($2000) for function f)one crate controller type A2 ($2000)plus a CAMAC crate and a CAMAC interface board inside the drivingcomputer which amounts to a total of $27000.In the new system, only 6 identical CPU modules are needed with a crateand a PC GPIB ($600) controller board, the whole totalling to $10000.

IMPLEMENTATION

To handle the CAMAC system, a library of subroutines was neededfor each kind of module. The implementing of such a software is compa-rable to the specific firmware for each application integrated in theCPU board of our new system. On the other hand, when a commercial IEEE488 device is available for a spectrometer (temperature regulation, DCsupplies etc..), its programming is condensed to a few simple instruc-tions - typically "write or read " character strings - instead of asérie of C(i)N(j)A(k)F(l) for a CAMAC module, if it exists.

EFFICIENCY

Several CAMAC modules (3 for "flipping" counting, 3 for P.S.D.time-of-flight etc...) can be replaced by a CPU board which providesthe automatic monitoring of all the microcommands of a function (e.g.switching of multiplexers for position reading, air pressure commandfor motor driving etc...). It is able to send and receive data in aformat adapted to a computer or a manipulator (e.g. formating of data,mnemonic language etc...). These characteristics have improved 10 to50 times the speed of execution of several existing devices (positionreading, transfer of PSD data etc..).

SYSTEM LIFECOST

The immediate financial advantages are further enhanced in oursystem by the use of the same kind of board (the CPU board) which con-siderably reduces maintenance duration, so efforts can be concentratedon the programming of a unique board. Besides, in the CAMAC system thedriving computer, being the intelligence of the system, is often deeplyinvolved in the BUS management; that is why the handling of the modulesis often "computer dependant" (specially with interrupt managementsystem) making the adaptation of the software on another type ofcomputer difficult. In our system, the peripherals are intelligentenough to be monitored by a simple non-computer dependant program andimprove the availability of the computer. Due to the universality ofIEEE 488 BUS, our peripherals can be used in other laboratories.

130

COMPOSITION OF THE NEW SYSTEM

The different components of our system were selected in relationto the performances required. Higher performances can be obtained withmore powerful material without modifying the described structure.

CONTROLLING MICROCOMPUTER

A widely commercialized product, the IBM PC XT or AT, controlseach spectrometer (this type of computer has practically become thestandard personal computer). It is relatively inexpensive ($4000-5000)and all improvements in PC's are made compatible with the IBM. Eachspectrometer can, therefore, be equipped with one or many microcom-puters, thus increasing their reliability and availability and mini-mizing the expenditure.

The microcomputer runs on MS DOS and has 512K RAM memory, one 20Mbytes disk drive, two 640K floppy diskettes, one graphics printer. Italso has a high resolution 600 X 400 pixels graphics monitor and acolor screen can be attached.

The controlling program itself is written in high level language.All the parameters defining the experiment are set up on one or manyfiles by means of interactive multiple windows software. To carry outan experiment, the monitoring program executes these run files. Duringthe acquisition period, the peripherals are autonomous, so the user canquit the controlling program to execute other programs (calculations,editing, image processing etc...).

The P.S.D. data image processing display depends on their size andshape. The XY data can be displayed in 3D under different angles andscales of color. The graphic software allows image processing likedisplay of several views, magnification of a window, cursormanipulation (eg. location of a given data). The data contained betweentwo amplitude marks can be extracted and displayed in another colorscale. Data are transmitted to the data processing computer by anetwork like ETHERNET.

THE MICROPROCESSOR BOARD

A commercial medium range CPU monoboard 8/16 bit, equipped with a6809 Motorola microprocessor was selected: the "DAFFODIL" board, manu-factured in 1983 by a Belgian firm.

It should be kept in mind that, within the structure of thedata acquisition and control system, this CPU board fulfills the roleof interface between "pure" electronics and the computer. Due to thepresence of RS232 and IEEE 488 components, it can be connected to anykind of computer. Five Versatile Interface Adapter (VIA) 1C's providethe CPU board the following programmable capabilities:- high performance digital I/O (80 Bidirectionnal I/O TTL lines) withmultiple "Handshake" and strobed I/O modes

- non standard serial interface- external pulse counting provided by nine 16-bit counter-timer- calibrated pulse generation- programmable frequency square waves generation

131

SOLENOID VALVE

RS232. TEftHMAL

RS232 LIHt

MICROCOMPUTERCONTROLLING

THE EXPERIMENT

CPU OAFFOOLWITH SOFTWARE

"MOTORCONTROL"

MONITORING BUS

b)CPU OAFFOC*.

WITH SOFTWARE"OPTICALENCODINGSYSTEM"

c)CPU OAF««.

WITH SOFTWARE"CONTROL OF

POWER SUPPUES-

IEEE 4IIBUS

MICROCOMPUTERFOR DATA

PROCESSINGJ

WITH SOFTWARE-FAST TIME OFFLIGHT FOR A

SMGLE COUNTER"

flcPU OAFFOOLWITH SOFTWARE

"DATAACaUISITION FORPSD WITH TIME

OF FLIGHTOPTION"

STEPPINGMOTOR SUPPLY

STEPPMGMOTOR

OPTICALABSOLUTEENCODER

MONITOR!« SUS

REGULATED DCPOWCT sum*

COUSOLENOC

CPU OAFFOOl.WITH SOFTWARE

MONITORINGOF NEUTRON

Old rfr*

t)

1

COUNTFR

1 1 MON1 I '""

CHOPPER

POSITIONSENSITIVE

DETECTOR (PS01

ANCILLARYIEEE DEVICES

COMMAND LEVEL CPU BOARDS LEVEL SPECTROMETER

Figure 2

132

The components of the DAFFODIL board enable it to undertake thefunctions expected from it, namely, easy connection to already existingequipments (fig n° 2) and conception of new functions.CPU BOARD FIRMWARE ORGANIZATION

The firmware consists of an on-board specific command interpreterfor each module (pulse counting, time of flight etc...) is written inmachine langage and uses a library of strict structured routines whichallow for quick and simple modifications and improvements. Thefirmware is mainly divided into 2 distinct categories: one covers theuser language (computer or manipulator) through IEEE 488 and RS232links and the other connects external electronics, or makes thefunction itself. Including on-site interactive testing of all thesubdivisions of the function (components, wires etc...), the firmwareof each module, generally, does not exceed 5 Kbytes half of which areidentical to all modules.

The programming of the CPU board is simple for two main reasons:- each board is the interface of a unique function or fulfills a uniquefunction, and then runs a few tasks (maximum 3)

- the programming of a new application does not interfere at all withthe existing ones.

The tasks run on different priority levels:1) low-level: command interpreter, orders execution and checking

external electronics to detect defects.2) medium-level: RS232 or IEEE 488 communication. This interrupt is

masked during the command completion.3) upper-level: interrupt management of VIAs request (start/stop,

counting overflow, "data present" etc...).

The in-board interpreter; its aim is to make the use of the moduleeasy to understand by non-specialists. The dialogue with thecontroller (computer or manipulator) consists in a character stringmade up of a mnemonic, followed by a variable number of parameters: eg"T1000" means start counting for 1000 1/100 sec.;"N100000" means startcounting for 100000 neutron pulses; "15,-10.123,7,3.25" means put-10.123 Amp. on power supply 5 and 3-25 Amp. on power supply 7-

To generate this language a few simple instructions are enough onthe driving computer:with a COMMODORE microcomputer in BASIC :

SET$="I5,-10.123,7,3.25" : PRINTsSUPPLY,SETSVarious command formats can be fed into the syntax analyser, but

the output format is adapted to the high-level languages on thecontrolling computer. The language syntax on all modules is identicaland a beginner can, with a general "help" command, use each modulewithout entering into the details of its functionning. A number ofcommands are reserved for the maintenance technicians to question theboard on its status and environment.

133

From the fonctionnai point of view, each character of the commandstring provokes a medium-level interrupt request until the validationterminator. This interrupt has, of course, no effect on the tasksrunning on upper-levels. The string is then entrusted to the syntaxanalyser which activates the corresponding processing subroutine.Connection with external electronics is provided by the judicious useof VIAs and their programming. Depending upon the problem to besolved, the corresponding operating mode is selected by writing intoVIA registers. Here again- a library of subroutines provides themanipulation of the more commonly used external I/O electronic systems,although the difference between the various sytems begins to appear.For example, in a DC supplies command board application, a certainprotocol respecting an established timing has been provided. In factit amounts to driving a specialized bus, defined by the conceptor ofthese DC supplies, and its addressing mode, encoding functions andsynchronization modes. The section of the program which undertakes thesupplies interfacing allows the DAFFODIL board to be the controller ofthis specific bus. Below is an example, in assembly language, of VIAprogramming to send data to an external device:

Writing of 10 digits with automatic "data ready" strobe

LDAfrOUTB ;all 8 bits of B port as outputSTA VIA-DDRB ;into Data Direction Register BLDAfrPSM ;Pulse ModeSTA VIA-PR ;into VIA Peripheral RegisterLDX#DATA ; "X" CPU register points on dataLDB#10 ;10 dataLOOP

LDA ,X+ ; reading dataANDA-fr$FO ;low order dataSTA VIA-PB ;each Port B output provokes "data ready"

; signalDECB

UNTIL EQ

Conception of one timer, four sealers and one preset sealer : each CPUboard offers 9 counters in its 5 VIAs. Four generate clock, the other5 are presetable can record 65536 pulses, each at a rate up to 600KHZ. An upper level interrupt is sent to the CPU when they overflow.The 24 bits amplitude is obtained by counting by software the numberof overflows. One clock is programmed to produce an interrupt every1/100 sec. The CPU counts these interrupts and detects either theright number of overflows of one sealer (preset sealer) or the presetnumber of clock interrupts (timer operation). In the "flipping" modewith polarized neutrons, another dissymetric programmable clock deter-mines the moments when magnetic field is to be set. At eachtransition of the clock, counters are fixed and their contentsmemorized during the stabilization of the field (1/100 sec). Thecounters ar'e then preset with the content they reached at the end ofthe preceding alternation and the cycle starts again. The end of thecounting is calculated with the preset value or the duration value whenthere is an even number of rotations.

134

Conception of time-of-flight acquisition with P.S.D.: The P.S.D.acquisition (formerly hardwired, conceived with specialized computers)is now totally fulfilled by the DAFFODIL board by software. All theacquisition parameters (eg. time channel number and its width) aresoftware selectable. Though the performances required are medium: 50KHZ counting rate;l6 or 24 bits amplitude per cell;256 time channels of5 \isec depth minimum, the essential difficulty lies in acquisition andcounting speed. The fact that the surface elements managed are notnecessarily single cells is another difficulty; they can be a set ofdifferent shaped cells grouped together.

The acquisition takes place in the following manner: the DAFFODILboard awaits the start analysis signal; once received through a sensi-tive VIA bit, it triggers a clock the period of which is defined by avalue written into a timer of another VIA. This period is the timechannel width. The CPU then awaits the "DATA PRESENT" signal from theP.S.D. hardware through another sensitive bit of a VIA that triggersthe latching of the data. This latching gives a buffer one data deep.Once received, the CPU simply reads the data, this reading sendsautomatically the "DATA TAKEN" signal. The cell number is combinedwith the running time channel number. Then the CPU finds out, throughan indirection map, to which set the cell hit belongs and, depending onthe time channel number, adds one in the definite address. Theindirection map's task consists in memorizing for each cell the addressof the first time channel of its set. This map is programmable cell bycell and/or in rectangular areas by means of the following command:"R7,1,8,16,10" means: the cells contained by the rectangle atintersection of column 1 and row 8 in bottom left corner, 16 cells wideand 10 cells high, belong to area "7".

To avoid losing time in saving and recalling CPU registers, thecorresponding programming sequence shown below does not run on aninterrupt level. Data transfer to the computer stops the acquisition,but only during its memory refresh - 50 milliseconds every 5 secondsfor a 32X32 cells P.S.D. which is quite negligible -.

Main acquisition sequenceWAIT LDA SIGNAL ;Waiting for the start acquisition signal

BNE WAITLDX PERIODSTX TIMER ;Triggering time channel clock

LOOP SYNC ; Waiting for the "DATA PRESENT SIGNAL"LDB CHANNEL ' ; Reading of the latched time channelLDX (CELL) ;Reading of the latched hit cell

;and finding its group setABXABX ; combining time and group setINC ,XLDA ENDCHANNEL ; Last channel ?BNE LOOP ; NoBRA WAIT ; Yes

The loop sequence runs in less than 20 microsec., and the VIAbuffer is freed 10 microsec after its reading.

135

CONCLUSION AND FUTURE DEVELOPMENTS

The implementation of the data acquisition and control system ofthe ORPHEE spectrometers, i.e. the basic counting modules, positionencoding, motor controlling and P.S.D. data acquisition took about 6months.

The delay of new instruments installation has been reduced fromone month to one week due to the autonomous parallel modular structure,the integrated on-site testing programs, and the system's coherence.Despite the diversity of instruments, only 10 different modulesfirmware are used in the entire facility and each spectrometer uses, onthe average 4-5 modules. The system is two or three times cheaper thanthe former CAMAC system, better adapted and easier to operate.

The costly motor driving and encoding systems on the experimentsare gradually being replaced by a peripheral, designed by theLaboratory. This system, which integrates the DAFFODIL board, is verymodular since the different types of position encoders: absoluteoptical, incremental, synchro/resolver (1/1000° accuracy) and poten-tiometer can be used with both D.C. and stepping motors.Acknowledgements

The author owes a great debt to P. FROMENT regarding both hismajor contribution to the design of the control system and his valuablehelp during its completion.

He also wishes to thank B. FARNOUX, R.J. PAPOULAR and S.M. SHAPIROfor very useful discussions concerning Neutron Spectrometry.

He is also much indebted to them for a critical reading of themanuscript as well as for helping him to cast this paper in its finalform.

DEDICATED CAMAC MODULESCAMAC CONTROLLER

POWER SUPPLYFOR

MOTORS

ANY EXTERNAL

ELECTRONICS

MOTOR

CAMAC style structure

136

POSITIONING DEVICE

IEEE 488RS485

COMPUTER

| CPU BOARD

A A.

! 11

POWER SU

FOR

MOTOR

PPLY

G

MOTOR

MOTOR

T REGULATION

CPU BOARD hANY INTERNAL

ELECTRONICS

(e.g. T° REGULATION)

T" PROBE

T° CONTROL

ANY OTHER DEVICE

I

I

Instrument approach style structure

137

VME COMMUNICATIONCONTROLLER

DEDICATED VME MODULES VME MODULES (E.G. CPU BOARDS)

ANY KIND OF COMMUNICATION

MOTOR

MOTOR——

FORMOTORS

-

ANY EXTERNAL

ELECTRONICS

VME style structure

THE COMPUTER

MOTHER BOARD OF

THE PC

OR EXTERNAL EXTENSION OF ITS BUS

VIDEO DISPLAY

KEYBOARD

PC ADD ON BOARDS

___

POWER SUPPLYFOR

MOTORS

ANY EXTERNAL

ELECTRONICS

MOTOR

PC with add on boards approach

138

Annex

TECHNICAL PROBLEMS IN ACQUISITION AND CONTROLSYSTEMS OF NEUTRONS SPECTROMETERS

Introduction

It is difficult to find a general solution for every case because the situation of each research center isdifferent. The hardware of the spectrometer is often very old, sometimes it is too modern and there is a problemof local maintenance because of a lack of trained staff. Very often there are no financial resources to update thesystems. Most of the time there is lack of organisation.

Because of this last point, the team (if there is one) in charge of the experiment does not have a clearview on the system architecture as a whole, and perhaps, does not even feel the need to have one. Yet, theadoption of a general architecture and standards is a key point. It will bring uniformity between the hardware andthe software components and as a consequence a better efficiency of the technical staff.

If a system is to be chosen what should be its qualities ?

The key factor is: simple enough to be mastered and to be improved by the local task force: It should be

. inexpensive

. based on easily available components

. easy to install and to customize

. very easy to test and to repair.

. etc...

There should be a good compromise between high technology and simplicity. Therefore it must

. be modular

. rely on widely accepted standards

. accept the connection of équipement from commercial vendors

. be able to reuse parts of a previous system

. etc...

OVERVIEW OF SOME COMMON ACQUISITION AND CONTROL SYSTEMS

This little presentation is of course not exhaustive. It is intentionaly kept simplified . Therefore it is notobjective and not corresponding with much more complex situations.

A> The CAMAC standard

Description:

The computer that controls the experiment is connected to a CAMAC crate. In this crate there areCAMAC modules that fulfill a specific task: for example, stepping motor control, position encoding and so on.(Sometimes the CAMAC controller is the computer itself, and sometimes inside the crate a CAMAC controlleris connected to the computer via a specific link)

Advantages

- The CAMAC is a widely used standard in Nuclear Research. Many different kinds of CAMAC modules forvarious needs can be found, either in the commercial market or developped by other research laboratories.

- CAMAC is a modular standard. Typically, if a new function is to be added, it is enough to insert a new modulein the crate.

- CAMAC is modular at the level of the communication between the CAMAC controller and the modules.Each module responds to a certain number of functions very well classified and defined for writing, reading andcontrol.

139

Inconveniences

- CAMAC is rather expensive because of its poor distribution in industrial firms

- It is rather difficult to make a CAMAC module for unskilled personnel

- Computer choice is limited since few computer makers offer CAMAC interface and firmware on theirproducts.

- CAMAC standard is not commonly used in industrial process control whereas systems can profit from industrialapplications (regulation, digitally controlled machine-tools etc...)

BWME or MULTIBUS type standard

Description

The computer is connected to a VME or MULTIBUS crate through standard communications links likeethernet, IEEE 488, RS232 etc... In the crate multiple processor boards are connected to specific modulesdedicated for each feature of the experiment. The processor boards can work together to fulfill a complicatedtask. They run under a real time operating system like OS 9. In fact the VME or MULTIBUS crate acts morelike a specialised computer than as a hardware extension of the monitoring computer whose role is mainly toprocess data.

Advantages

- VME standard tends to replace CAMAC: it is more and more often used in nuclear research

- there are now many kinds of VME modules available

- Very high acquisition rates can be reached by such a system

- Very complex and automated tasks can be performed by the system without control of the main computer

Inconveniences

- VME or MULTIBUS modules are still expensive

- VME modules are often overdimensioned for a simple use

- the knowledge necessary to master the BUS of the system is important

- the knowledge necessary to do the programming of each processor board is important. It is at the level of aUNIX like operating system.

- The boards inside the crate are not independent. In case of breakdown it may be difficult to debug the problem.

C) "Instrument" approach

Description

The computer is connected to several instruments or devices using a standard communication link likeIEEE 488 RS232 or RS485. Each of these devices fulfills a complete function like counting, positionning, PSDacquisition etc....The devices are completely independent from each other. They share only the link to thecomputer. From the computer point of view the programming is fairly simple: it consists of sending messages tothe devices and of receiving data using a more or less comprehensible syntax.

Advantages

- The system is conceptualy satisfying: for each function there is an independent and intelligent device

- Control and test of the devices are simple because they are independent, and physically separated

140

- IEEE 488 and RS232 are widely accepted standards and can be implemented easily in any kind of computer

- Many IEEE 488 and RS232 devices are available, connection of a new device is basically simple and direct

- No physical limitations of connecting the electronic boards on the same crate or of linking different crates

Inconveniences

- IEEE488 and RS232 devices are still expensive

- The design of a stand-alone device requires a good knowledge of how to write the firmware of amicroprocessor board.

D^ PC with add-on boards approach

Description

Add-on electronic boards are placed directly inside the PC. It is conceptualy like the CAMAC systemexcept that the PC BUS is used instead of the CAMAC BUS.

Advantages

- The PC BUS is maybe the most widely distributed standard

- Many kinds of boards for various needs can be found, often at a cheap price

- ADD-ON board come frequently with very good software. Often they can be programmed using simple BASICinstructions.

Inconveniences

- No computer choice: the electronic system is completely computer dependent

- Physical restraint to connect all the wires to the core of the computer

- Difficulty to design a PC board because it requires a very good knowledge of PC hardware system

141

SYSTEM PROPOSALS

A single system should be selected among these standards. And if it is not possible, not more than two.

Controlling computer

Due to its wide distribution and its increasing power the use of a PC is a good choice. It is inexpensiveand all improvements in PC's are made compatible. It is a general purpose computer, and it can be used also as adevelopment system for microprocessor boards. The development tools are also inexpensive. User-friendlyaspects of the program that controls the spectrometer can be met easily, using windowing software.

Communication with external devices

It is necessary to select a standard of communication between the computer and the external devices.RS232 and IEEE 488 can be opted for. They offer the possibility to implement easily new features to theexperiment by simply connecting other devices. IEEE488 BUS (as well as RS485) has the advantage overRS232 to allow multiple connections on the same cable. Many inexpensive IEEE 488 interface boards areavailable for a PC.

System structure

It seems that the "instrument Approach" is a good solution. Its main advantage is to be easy tounderstand because it is coherent with the organisation of the experiment into separate and functional parts, e.g.for a neutron spectrometer:

-counting-positioning-PSD acquisition-control of sample environnement-regulation- etc...

Because each of these functionnal parts are completely separated they can be designed carefully. Theycan also be tested easily which is very important in case of breakdown.

The major drawback to this approach is the difficulty to build such a stand-alone device.There are mainly 3 problems:

1) the device, if it is intelligent, requires a CPU board running with a specialized firmware2) the CPU board must be connected to a special hardware to fulfill the expected function3) the device must provide communication with the controlling computer.

However, one should keep in mind that there are several suppliers of stand-alone microprocessorboards providing REAL TIME BASIC like environnements. Consequently, there is no need to write drivers tosupport the communication with the computer since it is ready made. In addition, there is often no need foradditionnai hardware, the CPU board being directly able to fulfill the basic electronic functions like counting orTTLI/O.

In cases where the acquisition rates are not critical, which is normally the case for reactors with lowneutron fluxes, the many TIL I/O facilities used in combination with the on board components and with a highlevel language such as BASIC is a key point:

-it allows to reuse already existing electronic parts using, for example, a simple BCD connection.-it allows to connect directly a customized electronic board for special needs, opening the way to localdevelopment.

Conclusion

A structure for the acquisition and control system of spectrometers must be chosen. Once the choice hasbeen made, all the development effort must be done following the guidelines of the structure. If the aimedpurpose is the optimum exploitation of local ressources, the first step should be to adopt the most simple system.That does not necessarily means low end technology, in fact it is sometimes just the opposite: It must be takeninto account that due to the integration of the electronics, it is more and more feasible to design spectrometerinstrumentation. But it needs consistent training and information for local manpower.

142

LIST OF PARTICIPANTS

AUSTRIA

Badurek, G.

Rauch, H. (Observer)

CHINA

Ye Chuntang

FRANCE

Koskas, G.

Mason, S.A.

INDIA

Rao, K.R.

PAKISTAN

Butt, N.M.

PORTUGAL

Carvalho, F.G.

Margaça, F.A. (Observer)

SLOVENIA

Dimic, V.

Atominstitut,Schuettelstr. 115, A-1020 Vienna

Atominstitut,Schuettelstr. 115, A-1020 Vienna

Thermal Neutron Scattering Laboratory,China Institute of Atomic Energy,P.O. Box 275-30, Beijing 102413

CEA-CNRS,Laboratoire Leon Brillouin - CEN SaclayF-91191 Gif-sur-Yvette, Cedex

Diffraction Group, ILL,Institute Max von Laue - Paul Langevin,B.P. 156, F-38042 Grenoble

Solid State Physics Division,Bhabha Atomic Research Centre,Trombay, Bombay 400 086

PINSTECH,P.O. Nilore, Islamabad

Physics Department,LNETI/ICEN,Estrada Nacional 10, 2685 Sacavem

Physics Department,LNETI/ICEN,Estrada Nacional 10, 2685 Sacavem

Institut Jozef Stefan,Jamova 39,61111 Ljubljana

143

SWITZERLAND

Bauer, O.S. Paul Scherrer Institute,CH-5232 Villigen

UNITED STATES OF AMERICA

Rhyne, J. Research Reactor Facility,University of Missouri, Research Park,Columbia, Missouri 65211

INTERNATIONAL ATOMIC ENERGY AGENCY

Lewkowicz, I. Industrial Applications and Chemistry Section,Division of Physical and Chemical Sciences,International Atomic Energy Agency,P.O. Box 100, Wagramerstrasse 5,A-Vienna, Austria

Akhtar, K.M. Physics Section,(Scientific Secretary) Division of Physical and Chemical Sciences,

International Atomic Energy Agency,P.O. Box 100, Wagramerstrasse 5,A-Vienna, Austria

SM9U>O)

144

QUESTIONNAIRE ON IAEA-TECDOCs

It would greatly assist the International Atomic Energy Agency in its analysis of the effective-ness of its Technical Document programme if you could kindly answer the following questionsand return the form to the address shown below. Your co-operation is greatly appreciated.

Title: Use of neutron beams for low and medium flux research reactors:R&D programmes in materials science

Number: IAEA-TECDOC-836

1. How did you obtain this TECDOC?

[ ] From the IAEA:[ ] At own request[ ] Without request[ ] As participant at an IAEA meeting

[ ] From a professional colleague[ ] From library

2. How do you rate the content of the TECDOC?

[ ] Useful, includes information not found elsewhere[ ] Useful as a survey of the subject area[ ] Useful for reference[ ] Useful because of its international character[ ] Useful for training or study purposes[ ] Not very useful. If not, why not?

3. How do you become aware of the TECDOCs available from the IAEA?

[ ] From references in:[ ] IAEA publications[ ] Other publications

[ ] From IAEA meetings[ ] From IAEA newsletters[ ] By other means (please specify)[ ] If you find it difficult to obtain information on TECDOCs please tick this box

4. Do you make use of IAEA-TECDOCs?

[ ] Frequently[ ] Occasionally[ ] Rarely

5. Please state the institute (or country) in which you are working:

Please return to: R.F. KelleherHead, Publishing SectionInternational Atomic Energy AgencyP.O. Box 100Wagramerstrasse 5A-1400 Vienna, Austria

use of neutron beams for low and medium flux …iaea-tecdoc-836 use of neutron beams for low and...

Documents